![Chapter 1
Introduction: Ontologies, Semantic
Technologies, and Intelligence
Terry JANSSENa,1
, Leo OBRSTb
, Werner CEUSTERSc
a
Lockheed Martin Corporation, USA
b
The MITRE Corporation, c
State University of New York at Buffalo
Abstract: In recent years ontologies and semantic technologies more generally
have begun to be applied to assist the intelligence community, for information
integration, information-sharing, decision-support, and in many other applications.
This chapter introduces the topic of the book and provides background information
concerning its rationale, historical perspective, a vision for the future, and briefly
describes the chapters of the present volume.
Keywords: Ontology, information-sharing, intelligence community, semantic
technologies.
1. Why Ontologies: What do the Intelligence Community and Its Customers
Actually Need?
Probably the best way to address this question is look at what the Director of National
Intelligence (DNI) said in 2008 about the near future of intelligence, in the report titled
The DNI’s Vision 2015 [1]. We are unable to do justice to this insightful document in
this Introduction so the reader is encouraged to read it in its entirety. Some important
points are pulled from this report, not in perfect context, and are quoted here [1]:2
In this [adversarial, terrorism] environment [worldwide] the key to achieving
strategic advantage is the ability to rapidly and accurately anticipate and adapt to
complex challenges… [p. 6] By 2015 we will need integrated and collaborative
capabilities that can anticipate and rapidly respond to a wide array of threats and
risks… [p. 7] To succeed in this fast-paced, complex environment, the Intelligence
Community must change significantly. For example, our counterintelligence activities
face an array of new and traditional adversaries, yet we must operate within a
protected information-sharing environment that challenges existing notions of security
[and] of risk… [p.7] Analytic precision and accuracy will be merely the minimum
requirements expected by our customers; our accuracy must be clear, transparent,
objective and intellectually rich… they will expect instantaneous support ‘on demand’.
1
Corresponding Author: Terry Janssen, Lockheed Martin Corporation, IS&GS/GSS, Herndon, VA 20171,
USA; E-mail: terry.janssen@lmco.com.
2
Terms in bold are in the original document.
Ontologies and Semantic Technologies for Intelligence
L. Obrst et al. (Eds.)
IOS Press, 2010
© 2010 The authors and IOS Press. All rights reserved.
doi:10.3233/978-1-60750-581-5-1
1
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-13-320.jpg)
![By 2015, a globally networked Intelligence Enterprise will be essential to meet the
demands for greater forethought and improved strategic agility.
[But is important to note that] the purpose of intelligence is not merely to
determine truth, but to enable decision-makers to make better choices in dealing with
forces outside of their control. Intelligence helps reduce the degree of uncertainty and
risk when crucial choices are made. Our measure of success is simple: did our service
result in a real, measurable advantage to our side… [p. 10] [We] also need to exploit
commercial technologies to develop new ways of providing service. [p. 11] By 2015,
the Intelligence Community will be expected to provide more details about more issues
to customers. We anticipate different types of customers – with greater expectations –
and new demands to change the basic engagement model by which we serve them… [p.
11] To engage customers effectively, we must use sophisticated techniques… to elicit
‘What do you want to accomplish?’ … [and intelligence collection and analysis] will
become more of an relationship than an event… [p. 12] Our analytic products will
increasingly resemble customized services with an emphasis on maximum utility rather
than simple releasability. Under concepts such as effects-based analysis, we will
engage customers with ‘What-if?’ considerations in addition to ‘What?’ conclusions.
To do so, our analysts will leverage disparate data and analytic tools and services,
working in mission-focused distributed analytic networks… [p. 12] To respond to the
dynamic and complex threat environment of the 21st
century, our operating model must
emphasize mission integration – a networked knowledge sharing model that rapidly
pulls together disperse and diverse expertise and resources against specific missions…
[p. 13] [It will] require multiple integrated collection systems… [and a] fully
integrated processing, exploitation and dissemination architecture… [p. 14] There
will be more emphasis on multi-agency teams pursuing ‘multi-INT’ collection
strategies… We envision a collection community capable of rapidly fielding
technological innovations that contain needed information… Above all else is the
demand that the information reach those who need it, when they need it, in a form that
they can easily absorb. [p. 14]
The analytic community will be expected to understand and make judgments on a
broad spectrum of national security threats, support a more diverse customer set, and
cope with unprecedented amounts and types of information. Information overload
already presents a profound challenge… [and] the analytic community has no choice
but to pursue major breakthroughs in capability. Applying the principle of
Collaborative Analytics analysts will be freed to work in a fundamentally different way
– in distributed networks focused on a common mission. [p. 15] Information
overload will be averted through sophisticated data preparation and tools. In 2015,
new information will be tagged so tools can trace our data across our holdings.
Analysts will use such tools to mine the data, to test hypotheses, and to suggest
correlations. [p. 15] By 2015 the focus should shift from information sharing (e.g.,
interoperable systems, information discovery and access) to knowledge sharing [using
an automated approach to the extent possible, with the humans in the loop to
understand and present it accurately, and end-users to make decisions and act upon
that knowledge]. [p. 17] Although we will continue to rely on commercial best of breed
technologies and best practices, the Intelligence Community will still need to research,
development and field disruptive technologies to maintain a competitive advantage
over our adversaries. We cannot evolve into the next generation ‘S curve’
incrementally; we need a revolutionary approach… [p. 18] Creating a culture of
innovation will require greater focus on advanced concepts, technology and doctrine to
enhance leadership, organization alignment, and resources. [p. 19]
T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence
2
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-14-320.jpg)
![2. Ontologies: An Enabling Technology
Meeting the vision put forward by the DNI requires a fundamental shift in perspective
from data sharing to knowledge sharing. It takes no more than a moment’s reflection to
realize that knowledge cannot be shared unless it can be represented and communicated.
In other words, interoperation at the knowledge level means more than syntactic
interoperability and the sharing of data represented in standardized formats. The
receiving system must interpret the data in the manner intended by the sending system.
This means either that the communicating systems use common vocabularies with
agreed-upon meanings for the terms, or that the interchange be mediated by an
appropriate translation capability. It means that semantic information must be
explicitly represented in a form accessible to all parties to a communication, so that a
shared understanding of the knowledge being transmitted can be assured. This is
precisely the purpose served by formal ontology.
Ontologies represent the types of entities that exist in a domain, the relationships in
which they can participate, and the attributes of entities of different types. Publicly
available formal ontologies provide the basis for semantic interoperability by providing
standardized representations to define the semantics of knowledge being exchanged. It
is the thesis of this book that semantic technology used to address problems in the
intelligence community is one of the “disruptive technologies” needed to maintain our
competitive edge. Application of current-generation semantic technology can provide
an immediate benefit. In addition, the process of developing applications will inevitably
reveal important issues for which new research is needed, thus spurring advances in
technology that will result in further practical benefits.
3. A Resource: The National Center for Ontological Research (NCOR)
Recognizing the need for institutionalized leadership in semantic technology, the
National Center for Ontological Research (NCOR) [2] was founded in October 2005 as
a partnership of groups and institutions engaging in ontological research in the United
States, with the State University of New York at Buffalo and Stanford as principal
administrative sites. NCOR was established to serve as a vehicle to coordinate and to
enhance ontological research activities, with a special focus on the establishment of
tools and measures for quality assurance of ontologies, on training in and dissemination
of good practices in the ontology field, and on the creation of strategies to advance the
creation of federations of principles-based ontologies which work well together within
the hub-and-spokes framework of the sort currently being advanced within the US
Government’s Universal Core (UCore) and Command and Control (C2) Common Core
(C2Core) initiatives [3, 4].
In 2008 NCOR was contracted by the US Army Net-Centric Data Strategy Center
of Excellence to create a series of ontologies for use in the biometrics and C2
(command and control) domains, and also to create a standard Common Upper
Ontology based on BFO and DOLCE, for the representation of entities in real-world
domains. In 2009 NCOR worked with MITRE to develop UCore-SL, a Semantic Layer
for UCore 2.0, an XML-based vocabulary resource designed to support data sharing
sharing between agencies within the Department of Defense, Department of Homeland
Security, Department of Justice and Department of Energy.
T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 3
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-15-320.jpg)
![UCore and C2Core (and possibly other “common cores” in the future, depending
on need) are vocabularies that exist in the Community of Interest (COI) paradigm
within the federal government [5, 6]. In the COI paradigm, a modular architecture, as
depicted in Figure 1, acts as structure for the emerging range of vocabularies. UCore
spans all vocabularies and in particular immediately spans all common core
vocabularies such as C2Core. The common core vocabularies in turn span all
appropriate, COI vocabularies. COI vocabularies involve narrower domains and can be
hierarchically structured, as shown in the figure. A COI comes into existence when two
communities ascertain that they need to share information. The two communities
engage in a discussion of the kinds of data they have and wish to share, and the
vocabularies they use to refer to that data. Then they evolve an agreed upon vocabulary
for the data they wish to share, thus developing a specific COI vocabulary.
Figure 1. Universal Core, Common Cores, and COI Vocabularies
In addition some COIs will develop semantic models of their vocabularies, i.e.,
ontologies. Others will develop structural models in XML Schema. An example of an
ontology developed for a vocabulary is that of UCore-Semantic Layer (UCore-SL), an
ontology that provides a semantics for UCore [7, 8]. At this time, UCore-SL is not
officially part of UCore, but is a module under UCore Affiliates. Also, see [9, 10] for
an early advocacy for common Command and Control semantics.
When Figure 1 is compared with Figure 2 [11], a typical rendition of the layers of
ontologies, one notices that there is somewhat of a correspondence between the layered
vocabularies and the layered ontologies. However, in actuality, UCore addresses
objects that typically would reside in a mid-level or lower upper level ontology, i.e.,
person, organization, facility, location, etc., and their properties. It is expected that
UCore-SL would itself be embedded under an upper (sometimes called “foundational”)
ontology such as Basic Formal Ontology (BFO) [12, 13], which is discussed in more
detail below.
T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence
4
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-16-320.jpg)
![Figure 2. Ontology Architecture and Layers
4. A Resource: The Center of Excellence in Command, Control, Communications,
Computing and Intelligence (C4I Center)
C4I systems are essential to our national security. Recognizing the need for a strong
intellectual base for C4I, and the need for comprehensive educational programs in C4I,
George Mason University established the C4I Center in 1989 as the nation's first and
only civilian university-based entity offering a comprehensive academic and research
program in military applications of information technology. The Center performs
research and supports educational programs in a wide variety of C4I areas. A central
element in the C4I Center’s research is the formal representation of knowledge about
the military and intelligence domains in both interoperable and machine processable
forms. The Center is actively engaged in research to apply probabilistic ontologies to
predictive naval situation awareness, and is developing an open source probabilistic
ontology editor and reasoner. Another strong area of research is Battle Management
Language (BML), a formalism to support reasoning about military doctrine and
Command and Control (C2) processes, explicitly representing military task-based
operations [14].
5. An Ontology Success Story: The Peculiar History of the Gene Ontology
The most conspicuous successes in ontology technology thus far have been in the
biomedical field, and they result especially from the fact that the Gene Ontology (GO)
[15] has been so widely used as a resource for the integration of data in the domains of
molecular biology, bio-chemistry, functional genomics, proteomics, and related fields
now of increasing relevance to clinical research and treatment. It is noteworthy that, in
the period 2000-2007 there has occurred a 17-fold increase in use of the term
T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 5
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-17-320.jpg)
![‘ontology’ in the abstracts collected in the standard PubMed database of medical
literature, yet almost all of this increase is associated with references to the Gene
Ontology.
There exist over 11 million annotations relating gene products described in major
databases of molecular biology to terms in the GO. These annotations create linkages
between genes and proteins to specific types of biological phenomena. Data related to
some 180,000 genes have been manually annotated in this way, and the GO is hereby
making the results of divergent kinds of life science research comparable and
integratable.
The GO is a founding partner of NCOR, and NCOR has played an important role
in creating the OBO (Open Biomedical Ontologies) Foundry, a federation of the GO
and its sister ontologies used in biomedical research [16]. The OBO Foundry is now
serving as platform for the testing of NCOR strategies for quality assurance and
ontology integration. The goal of the Foundry initiative is to create the conditions under
which the data generated through biomedical research and clinical care will form a
growing pool, to which algorithmic techniques can be applied in ways which serve the
formulation and testing of clinical hypotheses at all levels. Efforts at ontology building
are still standardly conceived in pragmatic terms, as projects motivated by the need to
solve problems internal to the information technology needs of specific groups or
organizations. The Foundry, in contrast, reflects a view of ontologies which sees them
as lying outside the realm of software artifacts created to address specific local needs
and sees them rather as part and parcel of the scientific enterprise [17]. Ontologies are
from this perspective resources developed for the long term, freely available for use and
subject to constant criticism and update.
6. An Example Initiative: The Basic Formal Ontology
The hub, in the OBO Foundry and in a series of related initiatives, is Basic Formal
Ontology (BFO), a top-level ontology building on lessons learned from the develop-
ment of ontologies by logicians and philosophers over more than two millennia. BFO
was developed initially to support the work of experimental scientists but is now
increasingly gaining acceptance as a top-level ontology standard for general use.
When ontologies are developed, like database schemas, simply to address local
purposes, this will not only bring limited advantages in data integration but is indeed
likely to intensify the very problems of forking which ontologies were designed to
counteract. BFO is designed to serve as a top-level ontology standard that will constrain
the developers of domain ontologies in such a way as to work against these effects. It is
designed to be a very small, a true top-level ontology. This means that, in contrast to
the foundational ontologies DOLCE [18] and SUMO [19], with which it otherwise has
many features in common, BFO contains no terms which would properly belong within
the domains of those spoke ontologies which extend it. Use of BFO in a hub-and-
spokes framework thereby establishes a clear division of expertise. It provides both a
simple common starting point for scientists in creating their ontologies, and also a
common set of guidelines for ensuring that ontologies are thereafter developed and
maintained in tandem with each other. BFO and its associated guidelines for ontology
development have been refined on the basis of experience in application in the context
of the OBO Foundry, and they are now increasingly being used also outside the
scientific domain.
T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence
6
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-18-320.jpg)
![7. Training and Dissemination of Ontology Technology
In addressing its training and dissemination roles, NCOR has organized a series of
ontology tutorials and training workshops, working closely in this also with the
National Center for Biomedical Ontology. Jointly with the Ontolog Forum, the
National Institute for Standards and Technology (NIST), and the newly founded
International Association for Ontology and its Applications (IAOA) [20], NCOR
organizes the annual Ontology Summit, held at NIST in Gaithersburg, MD since 2005.
The theme for Ontology Summit 2010 is “Creating the Ontologists of the Future,” with
a focus on certification individual ontologists and accreditation of institutions for
courses of study for ontologists [21].
Jointly with JCOR, the Japanese Center for Ontological Research, NCOR
organizes the InterOntology (Interdisciplinary Ontology) conference series held in
Tokyo since 2006. Most importantly for our purposes, here, however, is the Ontology
for the Intelligence Community (OIC) conference, discussed below, which has become
a vibrant yearly forum for exchange of ideas on the role of semantic technology for
problems of interest to the intelligence community.
8. Ontology for the Intelligence Community (OIC): A Forum for Knowledge
Interchange
The first two OIC conferences were organized by NCOR Director Barry Smith, and
were held in Columbia, MD, in 2006 and 2007 [22] under the heading “Towards
Effective Exploitation and Integration of Intelligence Resources.” The third and fourth
OIC conferences were held in 2008 and 2009, respectively, at George Mason
University under the auspices of the George Mason University C4I Center [23, 24].
The OIC series was established to support the work of those who are using ontologies
to develop approaches to the analysis of intelligence that will enable greater flexibility,
precision, timeliness and automation of analysis and thereby maximize valuable human
resources in responding to fast-evolving threats. The OIC meetings have brought
together researchers and intelligence analysts from major agencies, universities and
other bodies involved in intelligence activities throughout the world. It provides an
important venue for exchange of ideas and sharing of insights on the role and effective
use of ontologies to problems in the intelligence domain.
Speakers at the first two OIC conferences included Werner Ceusters, Director of
the Ontology Research Group in the New York State Center of Excellence in
Bioinformatics and Life Sciences in Buffalo, who described how BFO-based ontologies
are being used to support the integration of instance data to enable tracking objects of
all kinds in computer representations; representatives from the FBI, NSA, CIA, other
USA agencies, the UK Defence Science and Technology Laboratory, and pivotal
technologists involved in applying ontology and semantic technologies to intelligence
needs; and Todd Hughes of the USA Defense Advanced Research Projects Agency
(DARPA). OIC 2008 invited speakers were Deborah McGuinness of Rensselaer
Polytechic Institute, one of the authors of the OWL Web Ontology Language; Michael
Gruninger of the University of Toronto, ontology researcher and active participant in
the ISO Common Logic standardization effort; and Leo Obrst, lead of the Information
Semantics Group at MITRE Corporation. The invited speakers at OIC 2009 were Chris
T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 7
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-19-320.jpg)
![Welty of IBM Watson Research Center, Doug Lenat of Cycorp, and a tutorial was
presented by Leo Obrst.
9. Semantic Interoperability for DoD and IC Systems: BML and Beyond
Modern military operations are increasingly becoming a joint endeavor, and it is hard to
imagine any situation in which a single service would operate alone. Instead, most real
world scenarios involve more than one country with multiple services, creating an
urgent need for international standards to support coalition interoperability. BML is
designed as an unambiguous language used to: 1) command and control forces and
equipment conducting military operations; and 2) to provide situational awareness and
a shared common operational picture. It can be seen as a standard representation of a
digitized commander's intent to be used for real troops, for simulated troops, and for
future robotic forces. BML is particularly relevant in a network centric environment for
enabling mutual understanding.
A BML development focus has always been conveying doctrinal knowledge
among military forces with diverse C2 processes. Thus, the advantages are clear of
evolving BML into a full ontology on military operations based on a strong formalism
for reasoning about tasks and actions. Such a C2 ontology would be a useful
complement to an intelligence community ontology.
Achieving semantic interoperability among DoD and IC systems is essential, but
one must recall that these systems will be always operating under the “fog of war”,
where incomplete data is the rule and uncertainty is ubiquitous. Unfortunately, current
ontology technologies provide no support for representing and reasoning in a principled
way with uncertainty and incomplete data. PR-OWL, a Bayesian first-order logic
extension to the ontology language OWL, was developed at the GMU C4I Center, and
is an example of current efforts to build probabilistic-aware ontologies [25]. These and
other similar efforts can be seen promising enabling technologies for the vision set
forth in the DNI’s Vision 2015.
10. The Contributions
There are a number of themes in the chapters of this book. Many of these themes span
multiple chapters, and many chapters have multiple themes. For example, many of
these chapters focus on supporting intelligence analysts using semantic technologies
and develop proofs of concept, especially the earlier chapters. The Ceusters and
Manzoor chapter, the Ulicny, Kokar, and Matheus chapter, Obrst et al, Lee, Shalliday
et al, Self et al, Ressler et al primarily address this theme. Another theme, however, is
that of dealing with uncertainty in ontology-based technologies, and hence addressing
the interaction between ontology and epistemology. Chapters that focus on this theme
include the Janssen et al, Laskey et al, and Huhns et al chapters. Finally, the first
chapter (the current chapter) and the book’s final chapter focus primarily on the impact
of ontologies and semantic technologies on intelligence collection and analysis – the
former attending to past and current efforts, the latter addressing issues about and
potential impacts on the future. The final chapter is that by Obrst, Ceusters, and Janssen.
T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence
8
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-20-320.jpg)
![situations, which can be combined to provide sets of possible situations that are
consistent with the evidence. The combination, according to the authors, provide
analysts with a way to gauge competing hypotheses and reduce the uncertainty of their
potential outcomes. Magellan uses XMLBIF (eXtensible Markup Language Bayesian
Interchange Format), along with OWL, RDF, and the SPARQL query language.
Chapter 11, “Geospatial Ontology Trade Study” (Ressler, Dean, Kolas) is focused
on the ontologies and other semantic standards that are valuable to geospatial analysis,
of particular interest to those involved with the GEOINT discipline, but also others who
require geospatial reasoning. Geospatial semantics ranges over representation of
geometry (i.e., points, lines, spaces, spheres, etc., involved in addressing regions),
geopolitics (i.e., borders, locations defined politically), temporal notions (how
geospatial notions change over time), and geographical and topographical knowledge.
As part of their trade study, the authors provide a matrixed view of the semantic
In the concluding chapter, Chapter 12: Ontologies, Semantic Technologies, and
technologies for addressing intelligence analysis and collection. A discussion is
presented on the use cases for the applications of these technologies vs. their
complexity, as gauged by the required expressiveness of the semantic models needed to
provide those applications. Cost must be addressed too, and measured against potential
benefits, as some emerging ontology cost models intend to provide. In addition,
emerging standards and technologies are discussed, from the SPARQL query language,
triple stores (that are repositories of OWL/RDF instance data, structured in graphs), and
rules and rule languages such as the Rule Interchange Format (RIF) – all of which are
supported by rapidly emerging tools. A prospective lesson is given for intelligence that
draws on the experience of using realist ontologies in biomedicine and healthcare, in
the hope that some notion of the value of ontologies and semantic technologies can be
indicated for intelligence collection and analysis. A distinction is then made between
ontology (the ways things are) and epistemology (the ways things are believed to be or
that we have current evidence for). Both technical disciplines and their tools are crucial
for intelligence, and provide complementary value. A human being has only one birth
date (ontology), but which of several ascribed to a particular person is correct
(epistemology)? Finally, the authors express optimism about the emerging convergence
of intelligence analysis and semantic technologies, and the potential value of that
convergence for intelligence.
REFERENCES
[1] Vision 2015: A Globally Networked and Integrated Intelligence Enterprise. Office of the Director of
National Intelligence (ODNI), July 22, 2008. http://www.dni.gov/reports/Vision_2015.pdf.
[2] National Center for Ontological Research (NCOR). http://ncor.us/.
[3] Universal Core (UCore). https://ucore.gov. [requires authorized login]
[4] Command and Control Common Core (C2Core). https://www.us.army.mil/suite/page/473883. [requires
authorized login]
[5] Communities of Interest (COI) – Home, Assistant Secretary of Defense (Networks and Information
Integration) Department of Defense Chief Information Officer (ASD(NII)/DoD CIO).
http://www.defenselink.mil/cio-nii/sites/coi/coi.shtml.htm.
described is that of the future and the potential promise of ontologies and semantic
standards that are possibly most applicable and important to geospatial analysis.
Intelligence Looking Toward the Future” (Obrst, Ceusters, Janssen), the view
:
“
T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 11
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-23-320.jpg)
![[6] DoD Directive 8320.2. Data Sharing in a Net-Centric Department of Defense. December 2, 2004,
Certified Current as of April 23, 2007. http://www.defenselink.mil/cio-
nii/sites/coi/Governance/832002p.pdf.
[7] Smith, Barry; Lowell Vizenor; and James Schoening. 2009. Universal Core Semantic Layer. In
Proceedings of the Ontologies for the Intelligence Community (OIC) conference, October 20-22, 2009,
George Mason University, Fairfax, VA.
[8] Smith, Barry. 2009. Universal Core Semantic Layer (UCore SL): An Ontology-Based Supporting Layer
for UCore 2.0, UCore Conference, MITRE, McLean, VA. September 23, 2009.
[9] Chaum, Erik; Richard Lee. Command and Control Common Semantic Core Required to Enable Net-
centric Operations. AFCEA-George Mason University Symposium "Critical Issues in C4I" May 20-21,
2008.
[10] Winters, Leslie and Andreas Tolk. 2009. C2 Domain Ontology within Our Lifetime. In Proceedings of
the 14th International Command and Control Research and Technology Symposium (ICCRTS), Jun 15-
17, 2009, Washington, DC.
[11] Semy, S.; Pulvermacher, M.; L. Obrst. 2005. Toward the Use of an Upper Ontology for U.S.
Government and U.S. Military Domains: An Evaluation. MITRE Technical Report, MTR
04B0000063,November 2005.
http://www.mitre.org/work/tech_papers/tech_papers_05/04_1175/index.html.
[12]. Basic Formal Ontology (BFO). http://www.ifomis.org/bfo.
[13] P Grenon, B Smith. 2004. SNAP and SPAN: Towards dynamic spatial ontology. Spatial Cognition
and Computation: An Interdisciplinary Journal, Vol. 4, No. 1. (2004), pp. 69-104.
[14] Battle Management Language (BML). http://c4i.gmu.edu/BML.php.
[15] Gene Ontology. http://www.geneontology.org.
[16] Open Biomedical Ontologies (OBO) Foundry. http://obofoundry.org.
[17] Smith, Barry. “Ontology (Science)”, in C. Eschenbach and M. Gruninger (eds.), Formal Ontology in
Information Systems. Proceedings of the Fifth International Conference (FOIS 2008), Amsterdam: IOS
Press, 21-35. http://precedings.nature.com/documents/2027/version/2.
[18] Descriptive Ontology for Linguistic and Cognitive Engineering . DOLCE. http://www.loa-
cnr.it/DOLCE.html.
[19] Suggested Upper Merged Ontology (SUMO): http://www.ontologyportal.org/.
[20] International Association for Ontology and its Applications (IAOA). http://www.iaoa.org/.
[21] Ontology Summit 2010: Creating the Ontologists of the Future. http://ontolog.cim3.net/cgi-
bin/wiki.pl?OntologySummit2010.
[22] Hornsby, Kathleen Stewart, ed. 2007. Proceedings of the Second International Ontology for the
Intelligence Community Conference (OIC) 2007, November 28-29, 2007, Columbia, MD, USA. CEUR
Workshop Proceedings, Volume-299. http://ftp.informatik.rwth-aachen.de/Publications/CEUR-
WS/Vol-299/.
[23] Laskey, Kathryn Blackmond; Duminda Wijesekera. 2008. Proceedings of the Third International
Ontology for the Intelligence Community Conference. Fairfax, VA, USA, December 3-4, 2008. CEUR
Workshop Proceedings Volume-440. http://sunsite.informatik.rwth-aachen.de/Publications/CEUR-
WS/Vol-440/.
[24] Costa, Paulo; Kathryn Laskey; Leo Obrst, eds. 2009. Proceedings of the 2009 International Conference
on Ontologies for the Intelligence Community. Fairfax, VA, USA, October 21-22, 2009. CEUR
Workshop Proceedings Volume-555. http://sunsite.informatik.rwth-aachen.de/Publications/CEUR-
WS/Vol-555/.
[25] Costa, Paulo; Kathryn Blackmond Laskey. 2006. PR-OWL: A Framework for Probabilistic Ontologies.
Proceedings of the Fourth International Conference on Formal Ontology in Information Systems.
November 2006. http://ite.gmu.edu/~klaskey/papers/FOIS2006_CostaLaskey.pdf.
T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence
12
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-24-320.jpg)
![Chapter 2
How to Track Absolutely Everything
Werner CEUSTERS1
, Shahid MANZOOR
Ontology Research Group, New York State Center of Excellence in Bioinformatics and
Life Sciences, University at Buffalo, USA
Abstract: The analysis of events prior to and during September 11 revealed that a
smooth execution of the intelligence process is hampered by inadequate
information sharing. This caused a rethinking of the intelligence process and a
transition towards a ‘Globally Networked and Integrated Intelligence Enterprise’
with the goal that more detailed, tagged, and, therefore, traceable, information will
reach those who need it, when they need it, and in a form that they can easily
absorb. We present the referent tracking paradigm and its implementation in
networks of referent tracking systems as an enabling technology to make this
vision come true. Referent tracking uses a system of singular and globally unique
identifiers to track not only entities and events in first-order reality, but also the
data and information elements that are created to describe such entities and events
in information systems. By doing so, it meets the requirements of the Nation’s
Information Sharing Strategy.
Keywords: referent tracking
1. Introduction
Intelligence, as defined by the Central Intelligence Agency (CIA), is ‘the information
our nation’s leaders need to keep our country safe’ [1]. This information is produced
by the US Intelligence Community (IC), i.e. the departments and agencies cooperating
to fulfil the goals of Executive Order 12333 which stipulates that ‘The United States
intelligence effort shall provide the President and the National Security Council with
the necessary information on which to base decisions concerning the conduct and
development of foreign, defense and economic policy, and the protection of United
States national interests from foreign security threats’ [2]. This is achieved through the
performance of what is called the ‘intelligence process’ which consists of five steps:
(1) the determination of the information requirements, (2) the collection of raw data,
(3) the processing of the raw data into forms that are more usable for intelligence
analysts or other consumers, (4) the integration, evaluation and analysis of the data in
order to generate reports satisfying the requirements, and (5) the dissemination of the
results to the appropriate level [3]. This last step, typically, leads to new information
requirements which initiate a new cycle of the intelligence process.
1
Corresponding Author: Werner Ceusters, Ontology Research Group, New York State Center of
Excellence in Bioinformatics and Life Sciences, 701 Ellicott street, Buffalo NY, 14203, USA; E-mail:
ceusters@buffalo.edu
Ontologies and Semantic Technologies for Intelligence
L. Obrst et al. (Eds.)
IOS Press, 2010
© 2010 The authors and IOS Press. All rights reserved.
doi:10.3233/978-1-60750-581-5-13
13
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-25-320.jpg)
![1.1. Challenges and barriers
Ideally, the information that is finally disseminated is (1) reliable, thus corresponding
faithfully to what is the case in reality, (2) complete, such that nothing what is essential
or required for the consumer to make adequate decisions is missing, (3) relevant, such
that decisions can be made efficiently, and (4) timely, guaranteeing that decisions can
be made early enough for the resulting actions to have the desirable effect.
Unfortunately, this ideal is very hard to achieve because of many barriers and
challenges [4]. A large number of these challenges are brought about by the
multiplicity of agencies, organizational levels within these agencies and information
consumers that are involved.
Although each step in the intelligence process comes with its own challenges, the
multiplicity of involved actors affects primarily the information requirements
assessment and the data-integration and analysis steps. So do the information
requirements that a specific organizational level has to take into account not only
consist of the external requirements put forward by the consumers to whom
intelligence reports of a specific nature and content need to be delivered, but also of the
internal requirements which determine what sorts of detailed information elements are
required and accessible to provide high quality reports. The integration, evaluation and
analysis step can be hampered by insufficient lower-level data (both quantitatively and
qualitatively), wrong information, and lack of meaningful data linkage. The net effect
is that the reliability, completeness and relevancy of the resulting conclusions suffer
considerably.
Although these three notions are intuitively straightforward, they can be defined in
various ways and for each such way, objective quantification is hard, if possible at all.
Furthermore, these notions are not entirely independent from each other. Reliability,
for instance, relates to accuracy which itself relates to relevancy: the more a
measurement is accurate, the more reliable it seems to be, yet, the relevancy of it might
diminish depending on the objectives of the intelligence effort: whereas providing
information on the duration of intercontinental flights in minutes to compare the
performance of foreign carriers with that of national ones seems reliable, accurate and
relevant, doing so in hours is hardly reliable, while in seconds for sure not relevant.
Redundancy of information elements within a collection of information will not harm
the completeness and relevancy of that collection as a whole, but for sure the relevancy
of the redundant elements themselves. At the other hand, from a second order
perspective, the presence of redundant information, if obtained from various
independent sources, might be an indication for the reliability of the collection.
1.2. Intelligence and Security Informatics
The analysis of events prior to and during September 11 revealed that a smooth
execution of the intelligence process is hampered by inadequate information sharing [5].
Not only are there legal and cultural barriers to information sharing – the ‘need-to-
know’ culture during the Cold War is now recognized to be a handicap in dealing with
terrorism and other asymmetric threats [6] – it is also technically very difficult to
integrate and combine data that are stored in different database systems running on
different hardware platforms and operating systems [7]. Although the Office of
Homeland Security, in 2002, identified information sharing across jurisdictional
boundaries of intelligence and security agencies as one of the key foundations for
W. Ceusters and S. Manzoor / How to Track Absolutely Everything
14
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-26-320.jpg)
![ensuring national security [8], the appropriate infrastructure is not yet there. This
recognition led to the development of a new science: ‘Intelligence and Security
Informatics’ (ISI) [9], which is commonly defined as ‘the study of the use and
development of advanced information technologies, systems, algorithms, and databases
for national- and homeland-security-related applications through an integrated
technological, organizational, and policy-based approach’ [10].
ISI tries to overcome the barrier that data which reside in distinct data sources are
organized in different schemas, and therefore are difficult to integrate. But still, once
some sort of integration has been achieved, it remains often very hard to determine, for
instance, whether two distinct pieces of information are about the same entity or which
piece of information is correct when several pieces about the same entity can’t be true
at the same time. As an example, a case study in a local police department revealed that
more than half of the suspects had either a deceptive or an erroneous counterpart
existing in the police system: 42% of the suspects had records alike due to various
types of unintentional errors, while about 30% had used intentionally a false identity
[11]. Deception is in the context of ISI a very hard problem indeed; it is not limited to
providing false identities, but includes also ‘cognitive hacking’ which involves
disinformation attacks on the mind of the end user of a networked computer system
such as a computer connected to the Internet [12]. Identifying such attacks is crucial in
an era in which the Intelligence Community seeks to make better use of Open Source
Information (OSINT) [13].
1.3. Vision 2015
To further advance the modernization of the information technology within the
Intelligence Community, the Office of the Director of National Intelligence [14]
published in February 2008 its ‘Information Sharing Strategy’ report [6], followed in
July by the ‘Vision 2015’ document [15]. They key idea, first introduced in the
National Intelligence Strategy [16], is the move towards a ‘Globally Networked and
Integrated Intelligence Enterprise’ with the goal that more detailed, tagged, and,
therefore, traceable, information will reach those who need it, when they need it, and in
a form that they can easily absorb. Efforts in these directions are expected to create the
ability to develop, digest, and manipulate vast and disparate data streams ‘about the
world as it is today’ by means of tags that enable the use of tools that can ‘trace related
data across our holdings, to mine the data, to test hypotheses and to suggest
correlations’ in addition to ‘measuring performance’ [15].
The key characteristics of the new information sharing model are [6]:
C1. ‘responsibility to provide’: sharing intelligence data while still addressing
the need to protect privacy, civil liberties, and sources and methods;
C2. enterprise-centric: providing services across agencies, partners, and
international borders for multiple mission use;
C3. mission-centric: able to adapt rapidly to changing needs and new partners;
C4. information-centric: security built into the data and environment using tags;
C5. attribute-based: access based on attributes that go beyond security
classification (e.g. environmental, affiliation, mission focus, etc.);
C6. data ‘stewardship’ (rather than data ‘ownership’), focusing on quality and
reusability of data rather than, but not excluding, protection.
W. Ceusters and S. Manzoor / How to Track Absolutely Everything 15
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-27-320.jpg)
![such as female, male, mosaic, etc., but that information of the precise gender of that
person is not available in that information system, or (2) that the gender of that person
is known to be of a type which scientifically has not yet been determined. Another
example is that if at a certain time the gender of a specific person is registered in some
information system as ‘male’, and at a later time as ‘female’, then there is, under
existing data storage paradigms, no way to derive from this change whether the change
in the information system reflects (1) a change in reality, for instance, because the
person underwent transgender surgery, (2) a change in what became known about
reality: the person’s gender might because of a congenital disorder not have been
determinable at the time of birth, but only later after several investigations, or (3) that
there was no change in reality or what we know about it, but that at the time of the first
entry a simple mistake was made. One can even imagine a fourth possibility, namely
that the meaning of the word ‘female’ would have been changed. The latter might seem
to be too far fetched – in fact, this did never happen for the words ‘male’ and ‘female’ –
but there are several examples in the past that come close. The title ‘Chief Executive
Officer’, for instance, was introduced in Europe in the late eighties, replacing titles
such as ‘Director General’ or ‘Managing Director’. A change in title, in those days, for
sure did not entail a change in position or power of the person to whom the new title
was attributed.
These types of issues are insufficiently addressed in modern Semantic Web
applications because they are not yet generally recognized: attempts to address them
are sparse.
2.2. Missing semantics
The most recent hype in information system networking is semantic interoperability.
By ‘semantic interoperability’, it is meant the ability of two or more computer systems
to exchange information and have the meaning of that information automatically
interpreted by the receiving system accurately enough to produce useful results, as
defined by the end users of both systems. Current attempts to achieve semantic
interoperability rely on agreements about the meaning of so-called concepts stored in
terminology-systems, such as nomenclatures, vocabularies, thesauri, or ontologies, the
idea being that if all computer systems use the same terminology, they can understand
each other perfectly. The reality is, however, that, rather than one such terminology
being generally adopted, the number of terminology-systems with mutually
incompatible definitions or non-resolvable overlap amongst concepts grows
exponentially, thereby contributing more to the problem of semantic non-
interoperability than solving it. Of course, ontologies developed for different purposes
can only reasonably be expected to have partial overlap, but more efforts should be
conducted to exploit overlap when resolvable.
3. Fundamentals of realism-based ontologies and data repositories
In contrast to traditional terminology approaches, the realist orientation in terminology
and ontology is based on the view that terms in terminologies are to be aligned not on
concepts but rather on entities in reality [17]. Central to this view are three assumptions
[18]. The first is that reality exists objectively in itself, i.e. independent of the
perceptions or beliefs of cognitive beings. Thus not only do a wide variety of entities
W. Ceusters and S. Manzoor / How to Track Absolutely Everything 17
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-29-320.jpg)
![exist in reality (human beings, terrorists, guns, attacks, countries, ...), but also how
these entities relate to each other (that human beings are citizens of countries, that in
most attacks guns are used, and so forth) is not a matter of agreements made by
scientists or database modellers but rather of objective fact.
The second assumption is that reality, including its structure, is accessible to us
and can be discovered: it is scientific research that allows human beings to find out
what entities exist and what relationships obtain between them. It is intelligence
analysis that allows analysts to find out which specific human beings are terrorists.
The third assumption is that an important aspect of the quality of an ontology or
terminology is determined by the degree to which the structure according to which the
terms are organized mimics the pre-existing structure of reality.
In the context of information systems, it means that an important aspect of the
quality of an information system is determined by the degree to which (1) its individual
representational units correspond to entities in reality, and (2) the structure according to
which these units are organized mimics the corresponding structure of reality.
3.1. Faithful representations
The above assumptions form the basis for distinguishing between three levels of reality
which have a role to play wherever ontologies are used as artifacts for annotation and
tagging, and wherever automated or semi-automated reasoning is required to be able to
deal with an overload of information, parts of which can be expected to be wrong.
Ontologies and data repositories for the intelligence community are no exception to this.
The three levels are [18]:
• Level 1: the (first-order) reality ‘in the field’: the persons that are tracked, the
events that are monitored, the users of the information system, and so forth;
• Level 2: the beliefs and cognitive representations of this reality embodied in
observations and interpretations on the part of observers, data collectors,
analysts and others;
• Level 3: the publicly accessible concretizations of such cognitive
representations in representational artifacts of various sorts, of which
ontologies, terminologies and data repositories are examples. Ontologies
contain typically representations for what is generic, thus representing entities
such as person, weapon, war, and so forth. Repositories cover what is specific,
thus holding representations for entities such as President George W. Bush Jr.,
the gun that killed John F. Kennedy, The Gulf War, etc.
In line with the theory of granular partitions [19] we argue that complex representations
should be composed in modular fashion of sub-representations built out of
representational units that are assumed to correspond to portions of reality (POR).
Some characteristics of the units in a representation created for intelligence purposes
are:
• each such unit is assumed by the authors of the representation to be veridical,
i.e. to conform to some relevant POR as conceived on the best understanding
(which may, of course, rest on errors). Thus if in a data repository a
representational unit standing proxy for a specific person is associated with
the name ‘George Bush’, then, under the realist paradigm, we assume that a
person with this name exists or has existed (that on the basis of the name only
W. Ceusters and S. Manzoor / How to Track Absolutely Everything
18
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-30-320.jpg)
![it cannot be determined which specific person is meant, does not make the unit
non-veridical);
• several units may correspond to the same POR by presenting different though
still veridical views or perspectives, for instance at different levels of
granularity (one thing may be described both as being brown and as reflecting
light of a certain wavelength, or one event as an event of administering and of
consuming drugs);
• what units are included in a representation depends on the purposes which the
representation is designed to serve.
3.2. Keeping track of changes
The real world is subject to constant change, and so also is our knowledge thereof. To
keep track of these two sets of changes, any representation concerning a relationship
between entities should be associated with at least the following pieces of information:
(P1) an index for the time period during which the relationship obtains, (P2) an index
for the time at which the representation is made, i.e. the time at which the relationship
is (believed to be) known, (P3) an index for the time that piece of information is made
available in the system, and (P4) an identifier standing proxy for the author of the
representation.
Keeping track of these various types of information makes it possible not only to
track reality faithfully from an individual analyst or agency perspective, but also to
preserve the knowledge about what was known by whom and at what time after
information which was residing originally in distinct systems becomes merged. It also
allows to assess whether information is disclosed in a timely fashion.
Suppose, for instance, that at time t10 it is known by analyst A1 that suspect S was
since t9 member of group G of possible terrorists, but that an entry to that effect in the
information system of his agency is made available not earlier than at t11. Thus
between t10 and t11, that information was not accessible. Furthermore, in reality, it
might be that S was already member of G at t5. That information might have been
known in another agency since t6, and made available at that time in their information
system. When the information in the two systems becomes merged, for instance after
the Vision 2015 situation becomes reality, it can still be assessed what was known at
each point in time in each agency.
4. Fundamentals of Referent Tracking
Referent Tracking (RT) is a paradigm for information management that is distinct from
other approaches in that each data element has to point to a portion of reality in a
number of predefined ways (Figure 1). It has been introduced in the context of
Electronic Health Record keeping [20], but its applicability is wider than that,
examples being digital rights management [21] and corporate memories [22].
By ‘portion of reality’ is meant any individual entity or configuration of entities
standing in some relation to each other. By ‘entity’ is meant anything that exists or has
existed in the past, whatever its nature. A ‘configuration’ is a portion of reality which is
not an entity in its own right. Whereas a specific person, his or her activities, the social
network he belongs to, the analyst examining information about that person, and that
W. Ceusters and S. Manzoor / How to Track Absolutely Everything 19
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-31-320.jpg)
![examination itself are each individual entities, the configuration that the activities of
this person are being monitored by an intelligence agency, or his or her being part of
that social network, is not. Another example of a configuration is the being of an
engine in a car. Both that car and that engine are entities, but the fact that that engine is
in that car, is not. If that engine would not be in the car, but, for instance be placed by a
mechanic outside the car for repair purposes, still the very same entities (the car and the
engine) would be involved, but there would be another configuration.
Within the RT paradigm, configurations are referred to by means of a data type
called a ‘RT-tuple’, whereas entities are represented by means of a data type called
‘representation’. Both data types come in several forms depending on the nature of the
portion of reality they carry information about (see section 6).
RT, through its data types, allows also for the drawing of an explicit distinction
made in Basic Formal Ontology (BFO) [23] between specific entities called
‘particulars’ from generic entities called ‘universals’. Particulars are specific and
unique entities, unique in the sense that they each occupy specific regions of space and
time, and that nothing other than a specific particular can be that particular. Examples
are concrete persons such as George W. Bush Jr. and George W. Bush’s heart. Some
particulars, such as each of four tanks in a specific squadron, may exactly look the
same, but they are still distinct particulars. One can be destroyed, while the other three
remain intact. For particulars of specific interest, such as persons, ships, and hurricanes,
proper names are used to mark the importance of their individual identity. For other
particulars, such as cars or pieces of complex equipment, serial numbers are used for
unique identification purposes.
Portion of Reality
Entity
Particular
Universal
Defined class
Representation
Non-referring
particular
Information bearer
Denotator
IUI
RT-tuple
corresponds-to
Configuration
represents
CUI UUI
denotes
denotes
is about
Representational unit
denotes
contains
Figure 1: Reality and representations
W. Ceusters and S. Manzoor / How to Track Absolutely Everything
20
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-32-320.jpg)
![4.2. Denotators
A denotator is a representational unit which denotes directly an entity in its entirety
without providing a description. An example of a denotator is the string ‘Bush’ in the
sentence ‘President Bush visited Europe several times’ when, whether or not known to
the reader of the sentence in question, the writer had in mind a particular Bush, whether
George Bush Jr. or George Bush Sr. The sentence itself is an information bearer
according to our terminology. Because many representations are built out of constituent
sub-representations as their parts, in the way in which paragraphs are built out of
sentences and sentences out of words, RT uses the data type called ‘representational
unit’ to represent such smallest part. Examples are: icons, names, simple word forms,
or the sorts of alphanumeric identifiers found in digital records. Note that many images
are not composite representations since they are not built out of smallest
representational units in the way in which molecules are built out of atoms (Pixels are
not representational units in the sense defined.) [18].
RT distinguishes explicitly and formally between three types of denotators,
referred to respectively as ‘IUI’, ‘UUI’ and ‘CUI’.
An IUI – abbreviation for ‘Instance Unique Identifier’ – is a denotator in the form
of a persistent, globally unique and singular identifier which denotes (or is believed to
denote) a particular and which is managed in a referent tracking system. A UUI – for
‘Universal Unique Identifier’ is a denotator which denotes a universal within the
context of a realism-based ontology. A CUI – abbreviation for ‘Concept Unique
Identifier’ – is a denotator for entities of a type that is commonly and ambiguously
called a ‘concept’ [17], but which in BFO is called a ‘defined class’, and defined as a
subset of the extension of a universal which is such that the members of this subset
exhibit an additional property which is (a) not shared by all instances of the universal,
and (b) also might be exhibited by particulars which are not instances of that universal.
5. Referent Tracking System
A referent tracking system (RTS) is a special kind of digital information system which
keeps track of (1) what is the case in reality and (2) what is expressed in other
information systems about what is believed to be the case in reality. It does this
unambiguously by means of the data types just sketched – in the first place resorting to
IUIs – using principles and methods that assure – modulo the occurrence of errors, the
resolution of which is also covered by the RT paradigm – that an IUI is (1) persistent
because once created in a RTS it is never deleted, (2) globally unique because an IUI
denotes only one entity within an RTS, and (3) singular because within an RTS, there is
only one IUI for a specific entity.
Figure 2 shows the various components of an RTS and how an RTS can be used in
association with external information systems and terminology (or ontology) servers.
The direction of the arrows depicted therein shows the processing of service requests,
the communication, however, being bi-directional to accommodate responses to the
requests.
W. Ceusters and S. Manzoor / How to Track Absolutely Everything
22
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-34-320.jpg)
![5.1. Components of a referent tracking system
Figure 2: Components of a referent tracking system
Referent Tracking Server (Peers)
Referent Tracking System
Referent Tracking Data Access
Server
External
Information
System
Reasoning Server
Referent Tracking System User Interface(s)
User
User
Terminology Server
Vocabulary Thesaurus Nomenclature Concept System
Realism-based
Ontology
or or
or
or
Referent Tracking Data
Store
RTS
Proxy
Peer
RTS
Server
Proxy
Peer
Internal Ontology
IUI
Component
An RTS includes at least four types of components: (1) one or more referent tracking
servers, (2) one or more referent tracking system user interfaces, (3) an RTS Proxy Peer,
and (4) an RTS Server Proxy Peer. The components execute on one or more processors,
computers or computing devices. Further, all of the components of an RTS can run on
one computing unit; one or more components can run on one computing unit, while
others run on one or more other computing units; or the components may be distributed
among various computing units.
Each referent tracking server includes a data access server [24], which manages
service requests coming from an RTS Proxy Peer or RTS Server Proxy Peer and which
performs data manipulation on the server’s main component: a referent tracking data
store thereby assisted by a reasoning server. The latter performs various sorts of
reasoning functions by combining data from the data store with information coming
from external terminology servers. The type of reasoning that can be performed
depends on whether the terminology server contains nomenclatures, vocabularies,
thesauri, and so forth. The referent tracking server comes also with an internal ontology
which is a repository dedicated, for instance, to store information obtained during the
initialization process, access control information about authorized users and usages,
and so forth. The referent tracking system user interfaces allow direct users of the RTS
to perform (1) a variety of management functions such as registering new external
information systems, configuring a referent tracking server, adding additional referent
W. Ceusters and S. Manzoor / How to Track Absolutely Everything 23
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-35-320.jpg)
![In order to be able to deal with terminology servers and the various sorts of
knowledge sources they offer (nomenclatures, thesauri, ontologies, ...), the RTS
includes a Reasoning API which helps in sending reasoning queries uniformly to
different terminology servers. The Reasoning API has an abstract class called
OntologyConnector, which provides an interface to the external terminology systems
by means of services. The interpretations of the OntologyConnector services are
specific to a particular terminology server; therefore, a separate implementation of the
OntologyConnector is required for each terminology server which is used to annotate
the particulars in the RTS.
Description logics are widely used for building ontologies. The reasoners for such
ontologies may take from 1 second to a day to compute inferences over the ontology
classes depending on their size and definitional complexity. Therefore, instead of
always directly communicating with the reasoners for each ontology when a specific
query is launched, the RTS is able to store these queries and the results that have been
returned by these reasoners as an inference graph in a database [24]. Thus, because the
execution time of the OntologyConnector services can range from milliseconds to
minutes depending on the query execution time in the external terminology system, the
OntologyConnector caches the results returned from these systems. The cache is stored,
for instance, in a RDBMS. During the execution of any of the OntologyConnector
services, it first searches in the cache.
6. Referent Tracking Data Elements: RT-tuples
RT-tuples, although all corresponding to portions of reality, come in various flavors
depending on the sort of information they contain.
6.1. A-tuples
A-tuples correspond to the assignment by some agent of an IUI to a particular. For the
typical case, that particular is a pure first-order entity such as a specific person or a
specific building about which information is to be stored in the RT system. However,
by storing tuples, the RT system itself acts as an agent that assigns IUIs to the tuples
itself. Indeed, for each insertion of an A-tuple, there is a corresponding insertion of a
D-tuple that contains information about the corresponding A-tuple. To prevent infinite
regress, the assignment of these IUIs does not involve the generation of an additional
A-tuple, but is implemented through the use of these tuple-IUIs as an internal
annotation to the tuple itself.
Three factors can be distinguished as structural elements involved in such an
assignment act: (1) the generation of the relevant alphanumeric string, (2) its
attachment to the relevant object, and (3) the publication of this attachment [20].
A-tuples are of the form < IUIp, IUIa, tap > where IUIp is the IUI of the particular in
question, IUIa is the IUI of the author of the assignment act, and tap is a time-stamp
indicating when the assignment was made.
6.2. D-tuples
In light of the need or desire to resolve mistakes [25], RT includes the use of D-tuples,
which are to be created whenever (1) a tuple other than a D-tuple is added to the RTS
W. Ceusters and S. Manzoor / How to Track Absolutely Everything 27
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-39-320.jpg)
![between the particular p and the universal u at time t, which is such that p stands to no
instance of u in the relationship r at t [26, 27].
This ontological relation can be expressed by means of a ‘PtoU(-) tuple’ which is a
lacks-counterpart of the PtoU-tuple and has the form <IUIa, ta, r, IUIo, IUIp, UUI, tr>,
expressing that the particular referred to by IUIa asserts at time ta that the relation r of
ontology IUIo does not obtain at time tr between the particular referred to by IUIp and
any of the instances of the universal UUI at time tr.
6.7. PtoN-tuples
Important particulars such as persons, ships, hurricanes, and so forth are often given
proper names which function as denotators in reality outside the context of a referent
tracking system. This sort of information is stored in an RTS by means of one or more
‘PtoN-tuples’ where ‘N’ stands for ‘name’. These tuples have the form < IUIa, ta, nt, n,
IUIp, tr , IUIc >, where
• IUIa is the IUI of the author asserting that n is a name of type nt used by IUIc
to denote IUIp;
• ta is a time-stamp indicating when the assertion was made;
• IUIc is the IUI for the particular that uses the name n (this can be a person, a
community of persons, an organization, an information system, ...);
• IUIp is the IUI referring to the particular which the author associates with n;
• n is the name which the author associates with IUIp;
• nt is the nametype (examples being first name, last name, nick name, social
security number, and so forth); and
• tr is a time-stamp representing a time at which the author considers the
association appropriate.
7. Discussion
7.1. Referent Tracking and action-oriented formalisms
RT, at first sight, might look similar to other approaches. For instance, the need to track
objects through time as they change, and to reason (and to have machines sometimes
reason) over information that describes such changes, is what motivated calculi such as
the situation calculus, the event calculus, and the fluent calculus, as well as some
Knowledge Representation and Reasoning Systems. These approaches seek an efficient
solution to the projection problem [28]: given an action theory that specifies the
preconditions and effects of actions (including sensing), and a knowledge base about
the initial state of the world, determine whether or not some condition holds after a
given sequence of actions has been performed [29].
The situation calculus is a logic formalism that was first introduced by John
McCarthy in 1963 [30] and since then underwent a few modifications [31]. The basic
elements of situation calculus are: (1) actions that can be performed in the world, (2)
fluents that describe the state of the world, each fluent thus being the representation of
some property, and (3) situations. McCarthy and Hayes considered a situation to be ‘a
complete state of the universe at an instant of time’ [32], a position which is also
maintained in fluent calculus [33], whereas others redefined situations as finite
W. Ceusters and S. Manzoor / How to Track Absolutely Everything
30
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-42-320.jpg)
![sequences of actions, thus a history of actions [31]. Event calculus does without
situations, and uses only actions and fluents, whereby the latter are functions – rather
than predicates as is the case in situation calculus – which can be used in predicates
such as HoldsAt to state at what time which fluents hold [34].
RT differs in substantial ways from these logical formalisms. First of all, the goal
of RT is not just to represent actions and changes, but all entities that exist in reality.
Furthermore, these sorts of logics focus on computational aspects, but do not provide
an integrated ontological characterization of entities such as actions, plans, and,
because of their four-dimensionalist nature, for sure not of objects. It has been shown
that it pays off to add more ontological rigor to formalisms such as situation calculus,
for instance by using it only as one component for causal reasoning within a more
elaborate, multi-component system [35].
RT, in contrast, is not in the first place a computational framework, but rather a
representational one anchored in the realist view adhered to in Basic Formal Ontology
(BFO) [23]. BFO distinguishes, for instance, continuants (such as George W. Bush)
from occurrents (such as George W. Bush’s life or his last trip from Washington to
New York). These distinctions, including BFO’s treatment of locations, positions and
location schemes, was deemed essential in building a robot navigation model on top of
situation calculus as embedded in Kuipers’ Spatial Semantic Hierarchy [36].
Relationships of the sort expressed by, for instance, RT’s PtoP- and PtoU-tuples hold
only during certain time-periods [37, 38], and when they hold is expressed in the
corresponding tuples themselves. In addition, PtoU-tuples express what universals a
particular instantiates, thus also whether the entity described is an action or an object.
Although no attempt has been made thus far, it seems plausible to assume that it is
possible to express part of an RT database in terms of situation or event calculus.
7.2. Facts versus beliefs
The requirements within RT that tuples must make direct and explicit reference to that
what they are about, and that this can only be done for entities that exist or have existed,
would seem to make it very difficult to represent uncertain, or possibly deceptive
knowledge. One can wonder if, for example, an intercepted communication contains
‘Cain will strike down Abel’ and it is believed that ‘Cain’ and ‘Abel’ are code words
for non-personal entities, whether this belief can be recorded in this system. Similar
questions can be asked about things in the future: isn't it important for a
representational framework to be able to state knowledge about future happenings and
entities that might not exist until the future, such as tomorrow’s sunset or Al-Qaeda’s
next attack?
It is here that the distinction between three levels of reality as discussed in section
3.1 and the assignment of IUIs to RT-tuples themselves play a role. If a PtoP-tuple to
which IUI-457 is assigned states that George W. Bush was president of the US in 2007,
then the latter is taken to be a representation of reality – which of course may be a
mistake – whereas IUI-457 is the proposition that the latter is the case. That this
proposition is entertained (or not) by a specific person can be expressed by additional
PtoP-tuples that relate the tuple in question to that person by referring also to adequate
belief-related relations or processes depending on what sort of ontology is used. As in
the case of action logics, RT itself does not come with a logic of beliefs, but from the
representations, so we believe, secondary representations in terms of a belief logic can
be generated.
W. Ceusters and S. Manzoor / How to Track Absolutely Everything 31
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-43-320.jpg)
![For entities in the future, RT offers the possibility to reserve IUIs, rather than to
assign IUIs [20]. Thus it is possible to assign an IUI to the plan to see and enjoy next
Sunday’s sunset, whereas the detailed RT representation of that plan itself would
contain a reserved IUI for that particular sunset.
7.3. Maintaining integrity
There are several challenges in maintaining the representational integrity of an RT
system, specifically with respect to the requirements that an IUI within an RTS should
denote only one entity, and that there is only one IUI for a specific entity. If, for
instance, one doesn’t know that ‘Usama bin Ladin’ and ‘Osama bin Laden’ denote the
same individual, how could one possibly know to relate both names to the IUI denoting
that individual? Here responsibility for faithful representation is shared between the
user and the user interface. Whereas the former must devote enough effort to find out in
each specific case what individual a name denotes, the latter, assisted by additional
applications, must make it possible to reduce the effort required. Term comparison
algorithms might be used to inform a user that a name similar to the one entered is
already registered. Triggers and alerts can be implemented to warn a user that distinct
individuals have the same name, and so forth. All this, however, does not guarantee
that the right decision will be made in every case, and errors will very likely occur. So
there have to be procedures to detect and correct mistakes. It is here that the D-tuples
play an important role [25].
Easy to solve, once detected, are mistakes in which a particular has been assigned
more than one IUI. In this case, only one of these IUIs would be used in future tuples,
whereas all tuples in which the other IUIs are used will be replaced by tuples in which
that one IUI will replace the redundant ones. This mechanism guarantees that it still
remains known that during some period in the past, information concerning one
particular was believed to be about two or more particulars.
More work would be required in the opposite case, i.e. when the same IUI is used
to denote distinct particulars. Here it might be necessary to perform a manual revision
of the tuples in which that is used.
To detect mistakes, the ontologies in whose terms RT-tuples are expressed can be
used to guide integrity-checking routines that run over the RTDB. Because, for
instance, persons (or any material continuant) cannot be at two distinct places in the
same time, the presence of RT-tuples in the RTS that suggest this to be the case,
indicates a mistake of the type ‘one IUI for distinct particulars’. Logically, because two
distinct material continuants cannot occupy the same spatial region, any collection of
RT-tuples representing that this would be the case must contain an error of the type
‘distinct IUIs for the same particular’.
7.4. RT and the Semantic Web
The various types of tuples enumerated in section 6 are expressible using standard
Semantic Web technologies, though with some additional formalisms implemented at
the data-base storage level. This is indeed the approach that has been taken in
implementing the system [24].
The Resource Description Framework (RDF) [39] was used as the basic
representation language. Our RDF representations of the RT-tuples are treated as
resources themselves: each resource is therefore prefixed with the RTS name space
W. Ceusters and S. Manzoor / How to Track Absolutely Everything
32
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-44-320.jpg)
![URI and the prefix ‘rts:’ such that, for instance, the resource rts:IUI-1 is the same as
http://org.buffalo.edu/RTS#IUI-1. To declare properties for resources, we used RDFS
and mapped the RT-tuples to RDFS classes, thereby ensuring that the class names are
identical to the template names, with the exception of PtoU-, which, because of
restrictions in the RDFS naming conventions, has been mapped to PtoLackU.
Our implementation of the RTS is accessible through Web services which are
invoked through SOAP messages [40] containing the procedure information (procedure
name, parameters and return type) and port type (location of the procedure). The RTS
uses Axis for Java [41] to host the web services thereby taking advantage of the native
support of the Web Services Definition Language (WSDL) [42] that Axis provides.
The RTS has been build to be independent of any data source technology. To
achieve this goal, we have defined the RTRepository class as an abstract Java class.
This class provides all necessary services for managing the data based on the principles
defined in the RT paradigm. To manage the RT data in a specific data source
technology, an extension of the RTRepository for that specific technology is required.
We have decided to develop the RTRepositorySesameImp class by extending the
RTRepository such that it targets the SAIL Sesame API for manipulating RDF graphs
as a data source [43].
Because the RT data are expressed in RDF, RDF query languages such as RQL
[44], SPARQL [45] and SeRQL [43] can be used for retrieval. To this end, the
RTRepository comes with the service ‘repository.query(querystring, language)’ which
has an argument for the query string and a second one for the name of the query
language in which the first argument is expressed. The SeRQL query language is
implemented with the help of the Sesame SeRQL query language module, and the
SPARQL query language is implemented with the help of the ARQ query module (a
SPARQL processor for Jena) [46].
8. Conclusion: meeting the new intelligence criteria
When set up in appropriate ways, a network of referent tracking systems is able to meet
all the requirements identified for the envisioned Globally Networked and Integrated
Intelligence Enterprise (see section 1.3).
The requirement to share intelligence data while still addressing the need to protect
privacy, civil liberties, and sources and methods (C1), can be met by using the IUIs,
typically the ones that stand proxy for persons, as pseudonyms. It would even be
possible to go much further, for instance that all the information collected by credit
card companies, banks, department stores, telecom providers and so forth would be
pooled. Most citizens would find it unacceptable if that information were used for
intelligence purposes without there being any reason to do so. But with the appropriate
setup of IUIRepositories and RTDBs in such a way that, for instance, one specific
agency has the means to link IUIs to persons, but otherwise no access to other RT-data,
while other agencies would be able to do data-mining and pattern analysis on the
pseudonymized data, no privacy or civil liberties would be violated. When analysts
would detect suspicious patterns in the pseudonymized data pool, similar mechanisms
as search warrants can be used to obtain re-identification of the data.
The requirements to provide services across agencies, partners, and international
borders for multiple mission use (C2) and to be able to adapt rapidly to changing needs
W. Ceusters and S. Manzoor / How to Track Absolutely Everything 33
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-45-320.jpg)
![and new partners (C3) are supported by the possibility for referent tracking systems to
cooperate in growing networks.
The C4 requirement, i.e. to have security built into the data and environment using
tags, together with the C5-requirement that access should be based on attributes that go
beyond security classification, is met by the specific ways in which RT-tuples are set
up: they contain in every case an indicator for the provenance of the data and all data
are coded by means of ontologies or terminologies. Furthermore, each RT-tuple can be
treated as a first-order entity, thereby receiving its own IUI, and that IUI can be used in
other RT-tuples, for instance to describe to what type of entities or specific entities it
may be disclosed. The same IUI can be used to track the flow of the data-element
throughout the intelligence network.
Data stewardship, finally, focusing on quality and reusability of data rather than,
but not excluding, protection (C6) is a natural feature of the paradigm. One reason are
the principles for IUI assignment which require that before an IUI is assigned to an
entity, it should be checked whether that entity has already an IUI assigned to it.
Mistakes will happen, of course, but they are traceable over time; if, for instance, when
data accumulate, two IUIs start to appear repeatedly in the same configuration, then
they may stand proxy for the same entity. Or, if the database at some stage contains a
PtoP-tuple stating that the entity with IUIx was in some place at a given point in time,
while in a completely different place a bit later, then it is likely, modulo other types of
mistakes, that IUIx is denoting different things.
A problem, at first sight, might be the amount of work required to represent
information in this way. But here again, other types of software such as natural
language processing applications, might assist. Furthermore, as shown in [47, 48], it is
in many cases possible to translate structured information into a form that is RT-
compatible automatically. We argue that the effort to make systems of this kind
acceptable is not greater than the effort to bring about the change in mindset to realize
Vision 2015.
9. References
[1] Central Intelligence Agency. What is Intelligence? 2007 June 20, 2008 [cited 2008 August 12];
Available from: https://www.cia.gov/news-information/featured-story-archive/2007-featured-story-
archive/what-is-intelligence.html
[2] Reagan R. Executive Order 12333--United States intelligence activities. 1981.
[3] United States Intelligence Community. The Intelligence Process. 2008 [cited 2008 August 12];
Available from: http://www.intelligence.gov/2-business.shtml
[4] Travers R. A Blueprint for Survival; The Coming Intelligence Failure. Studies in Intelligence.
1997;Semiannual Edition, No. 1:35-43.
[5] Chen H. Intelligence and Security Informatics for International Security. Information Sharing and Data
Mining. New York: Springer-Verlag 2006.
[6] Office of the Director of National Intelligence. United States Intelligence Community Information
Sharing Strategy. 2008.
[7] Hasselbring W. Information system integration. Communications of the ACM. 2000;43(6):33-8.
[8] Office of Homeland Security. National Strategy for Homeland Security. Washington D.C.: Office of
Homeland Security 2002.
[9] H. Chen, R. Miranda, D. Zeng, T. Madhusudan, C. Demchak, Schroeder J. Intelligence and Security
Informatics: Proceedings of the First Symposium on Intelligence and Security Informatics (ISI’03).
New York: Springer-Verlag 2003.
[10] Chen H, Wang F-Y, Zeng D. Intelligence and Security Informatics for Homeland Security: Information,
Communication, and Transportation. IEEE Transactions on Intelligent Transportation Systems. 2004
December;5(4):329-41.
W. Ceusters and S. Manzoor / How to Track Absolutely Everything
34
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-46-320.jpg)
![[11] Wang GA, Atabakhsh H, Petersen T, Chen H. Discovering Identity Problems: A Case Study. IEEE
international conference on intelligence and security informatics. Atlanta, GA, USA Springer 2005.
[12] Cybenko G, Giani A, Thompson: P. Cognitive Hacking: A Battle for the Mind. IEEE Computer.
2002;35(8):50-6.
[13] Mercado S. A Venerable Source in a New Era: Sailing the Sea of OSINT in the Information Age. CIA
Studies in Intelligence. 2004;48(3):45-55.
[14] Office of the Director of National Intelligence. Office of the Director of National Intelligence. 2008
August 22 [cited 2008 August 28]; Available from: http://www.dni.gov/index.html
[15] Office of the Director of National Intelligence. Vision 2015: A Globally Networked and Integrated
Intelligence Enterprise. 2008.
[16] Office of the Director of National Intelligence. The National Intelligence Strategy of The United States
of America. Office of the Director of National Intelligence 2005.
[17] Smith B. Beyond concepts: ontology as reality representation. Proceedings of the third international
conference on formal ontology in information systems (FOIS 2004). Amsterdam: IOS Press 2004:73-84.
[18] Smith B, Kusnierczyk W, Schober D, Ceusters W. Towards a Reference Terminology for Ontology
Research and Development in the Biomedical Domain. KR-MED 2006, Biomedical Ontology in Action.
Baltimore MD, USA 2006.
[19] Bittner T, Smith B. A Theory of Granular Partitions. In: Duckham M, Goodchild MF, Worboy MF, eds.
Foundations of Geographic Information Science. London: Taylor & Francis Books 2003:117-51.
[20] Ceusters W, Smith B. Referent Tracking in Electronic Healthcare Records. In: Engelbrecht R,
Geissbuhler A, Lovis C, Mihalas G, eds. Connecting Medical Informatics and Bio-Informatics Medical
Informatics Europe 2005. Amsterdam: IOS Press 2005:71-6.
[21] Ceusters W, Smith B. Referent Tracking for Digital Rights Management. International Journal of
Metadata, Semantics and Ontologies. 2007;2(1):45-53.
[22] Ceusters W, Smith B. Referent Tracking for Corporate Memories. In: Rittgen P, ed. Handbook of
Ontologies for Business Interaction. New York and London: Idea Group Publishing 2007:34-46.
[23] Grenon P, Smith B, Goldberg L. Biodynamic Ontology: Applying BFO in the Biomedical Domain. In:
Pisanelli DM, ed. Ontologies in Medicine. Amsterdam: IOS Press 2004:20-38.
[24] Manzoor S, Ceusters W, Rudnicki R. Implementation of a Referent Tracking System. International
Journal of Healthcare Information Systems and Informatics. 2007;2(4):41-58.
[25] Ceusters W. Dealing with Mistakes in a Referent Tracking System. In: KS H, ed. Proceedings of
Ontology for the Intelligence Community 2007 (OIC-2007). Columbia MA 2007:5-8.
[26] Ceusters W, Elkin P, Smith B. Referent Tracking: The Problem of Negative Findings. In: Hasman A,
Haux R, Lei Jvd, Clercq ED, Roger-France F, eds. Studies in Health Technology and Informatics
Ubiquity: Technologies for Better Health in Aging Societies - Proceedings of MIE2006. Amsterdam:
IOS Press 2006:741-6.
[27] Ceusters W, Elkin P, Smith B. Negative Findings in Electronic Health Records and Biomedical
Ontologies: A Realist Approach. International Journal of Medical Informatics. 2007 March;76:326-33.
[28] Reiter R. Knowledge in Action. Logical Foundations for Specifying and Implementing Dynamical
Systems. Boston: MIT Press 2001.
[29] Vassos S, Levesque H. Progression of Situation Calculus Action Theories with Incomplete Information.
In: Veloso M, ed. Proceedings of IJCAI-07 2007.
[30] McCarthy J. Situations, actions and causal laws. Stanford, CA: Stanford University Artificial
Intelligence Laboratory; 1963.
[31] Reiter R. The frame problem in the situation calculus: a simple solution (sometimes) and a
completeness result for goal regression. In: Lifshitz V, ed. Artificial intelligence and mathematical
theory of computation: papers in honour of John McCarthy. San Diego, CA, USA: Academic Press
Professional, Inc 1991:359-80.
[32] McCarthy J, Hayes PJ. Some philosophical problems from the standpoint of artificial intelligence.
Machine Intelligence. 1969;4:463-502.
[33] Thielscher M. Introduction to the Fluent Calculus. Electronic Transactions on Artificial Intelligence.
1998;2(3-4):179-92.
[34] Kowalski R. Database updates in the event calculus. Journal of Logic Programming. 1992;12(1-2):121-
46.
[35] Kuipers B. The spatial semantic hierarchy. Artificial Intelligence. 2000 May;119(1-2):191 - 233.
[36] Bateman J, Farrar S. Modelling Models of Robot Navigation Using Formal Spatial Ontology. Spatial
Cognition IV Reasoning, Action, and Interaction. Berlin / Heidelberg: Springer 2005:366-89.
[37] Smith B, Ceusters W, Klagges B, Köhler J, Kumar A, Lomax J, et al. Relations in biomedical
ontologies. Genome Biology. 2005;6(5):R46.
[38] Smith B, Grenon P. The Cornucopia of Formal-Ontological Relations. Dialectica. 2004;58(3):279-96.
[39] Manola F, Miller E. RDF Primer. 2004 [cited; Available from: http://www.w3.org/TR/rdf-primer/
W. Ceusters and S. Manzoor / How to Track Absolutely Everything 35
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-47-320.jpg)
![[40] Mitra N. SOAP Version 1.2 Part 0: Primer. W3C Recommendation 2003.
[41] The Apache Software Foundation. Axis: A Webservices toolkit. 2005 [cited 25 January, 2007];
Available from: http://ws.apache.org/axis/
[42] Christensen E, Curbera F, Meredith G, Weerawarana S. Web Services Description Language (WSDL)
1.1. W3C Note 2001.
[43] Broekstra J, Kampman A, Harmelen Fv. Sesame: A Generic Architecture for Storing and Querying
RDF and RDF Schema. Lecture Notes in Computer Science - International Semantic Web Conference
ISWC2002. Heidelberg: Springer 2002:54-68.
[44] Foundation for Research and Technology – Hellas. The RDF Query Language (RQL). 2003 July 18
[cited 25 January 2007]; Available from: http://139.91.183.30:9090/RDF/RQL/
[45] Prud'hommeaux E, Seaborne A. SPARQL Query Language for RDF. W3C Working Draft 2006
October 4th [cited January 22, 2007]; Available from: http://www.w3.org/TR/rdf-sparql-query/
[46] RDF Data Access Working Group. ARQ - A SPARQL Processor for Jena. 2007 [cited 15th Febuary,
2007]; Available from: http://jena.sourceforge.net/ARQ/
[47] Rudnicki R, Ceusters W, Manzoor S, Smith B. What Particulars are Referred to in EHR Data? A Case
Study in Integrating Referent Tracking into an Electronic Health Record Application. In: Teich JM,
Suermondt J, C H, eds. American Medical Informatics Association 2007 Annual Symposium
Proceedings, Biomedical and Health Informatics: From Foundations to Applications to Policy.
Chicago, IL 2007:630-4.
[48] Manzoor S, Ceusters W, Rudnicki R. A Middleware Approach to Integrate Referent Tracking in EHR
Systems. In: Teich JM, Suermondt J, C H, eds. Proceedings of the American Medical Informatics
Association 2007 Annual Symposium Biomedical and Health Informatics: From Foundations to
Applications to Policy. Chicago IL 2007:503-7.
W. Ceusters and S. Manzoor / How to Track Absolutely Everything
36
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-48-320.jpg)
![Chapter 3
Uses of Ontologies in Open Source Blog
Mining
Brian Ulicnya
, Mieczyslaw M. Kokara,b
, Christopher J. Matheusa
a
VIStology, Inc.
b
Northeastern University
Abstract: The blogosphere provides a novel window into an important segment of
public opinion, but its dynamic nature makes it an elusive medium to analyze and
interpret in the aggregate, where it is most informative. We are developing a new
open-source blog mining technology that employs ontologies to solve this problem
by fusing the signals of the blogosphere and zeroing in on issues that are most
likely to migrate offline. This technology is designed to enable analysts to
anticipate the threats or opportunities these issues represent in a timely and
efficient fashion.
Keywords: Blog mining, Malaysia, human terrain, situation awareness
Introduction
Although much, perhaps even the majority, of what is discussed in the blogosphere is
of little consequence and fleeting interest, blogs continue to emerge as powerful
organizing mechanisms, giving momentum to ideas that shape public opinion and
influence behavior. There are nearly 16 million active blogs [17] on the Internet with
more launched every day, and bloggers have increasingly made an impact politically in
a range of places and situations. For example, Malaysian bloggers have recently
become quite assertive in confronting perceived corruption in their national
government despite strict governmental control of the major media [1]. Although one
must be careful not to extrapolate from the population of bloggers to the population as
a whole, clearly blogs provide unparalleled access to an important segment of public
opinion about events of the day.
The perspective blog mining provides is much more complete than that provided
by the letters to the editor section of a newspaper or magazine, if it has one. Even
premier print newspapers such as the New York Times publish only 15 to 20 of the
1000 letters they receive daily in reaction to their reports [5]; by contrast, there are
typically over 3,000 blog posts that cite New York Times stories for any particular day,
including many posts not in English. By mining the unfiltered reactions of bloggers to
the day’s events, we can clearly evaluate and quantify the reaction to the days’ news
reports of a highly motivated group of users. This is particularly useful where the local
press is tightly controlled.
Ontologies and Semantic Technologies for Intelligence
L. Obrst et al. (Eds.)
IOS Press, 2010
© 2010 The authors and IOS Press. All rights reserved.
doi:10.3233/978-1-60750-581-5-37
37
Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook
Copyright
©
2010.
IOS
Press,
Incorporated.
All
rights
reserved.](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-49-320.jpg)
![This ebook is for the use of anyone anywhere in the United States
and most other parts of the world at no cost and with almost no
restrictions whatsoever. You may copy it, give it away or re-use it
under the terms of the Project Gutenberg License included with this
ebook or online at www.gutenberg.org. If you are not located in the
United States, you will have to check the laws of the country where
you are located before using this eBook.
Title: Manette Salomon
Author: Edmond de Goncourt
Jules de Goncourt
Release date: September 11, 2020 [eBook #63179]
Most recently updated: October 18, 2024
Language: French
Credits: Produced by Carlo Traverso, Laurent Vogel and the
Distributed Proofreading team at DP-test Italia. (This
file was produced from images generously made available
by The Internet Archive/Canadian Libraries.)
*** START OF THE PROJECT GUTENBERG EBOOK MANETTE
SALOMON ***](https://image.slidesharecdn.com/8744-250407095250-687180fb/85/Ontologies-and-Semantic-Technologies-for-Intelligence-1st-Edition-L-Obrst-55-320.jpg)
Ontologies and Semantic Technologies for Intelligence 1st Edition L. Obrst
- 1. Ontologies and Semantic Technologies for Intelligence 1st Edition L. Obrst pdf download https://ebookgate.com/product/ontologies-and-semantic- technologies-for-intelligence-1st-edition-l-obrst/ Get Instant Ebook Downloads – Browse at https://ebookgate.com
- 2. Instant digital products (PDF, ePub, MOBI) available Download now and explore formats that suit you... Intelligence Support Systems Technologies for Lawful Intercepts 1st Edition Paul Hoffmann https://ebookgate.com/product/intelligence-support-systems- technologies-for-lawful-intercepts-1st-edition-paul-hoffmann/ ebookgate.com Social Web Evolution Integrating Semantic Applications and Web 2 0 Technologies Advances in Semantic Web and Information Systems 1st Edition Miltiadis D. Lytras https://ebookgate.com/product/social-web-evolution-integrating- semantic-applications-and-web-2-0-technologies-advances-in-semantic- web-and-information-systems-1st-edition-miltiadis-d-lytras/ ebookgate.com Semantic Web Technologies Trends and Research in Ontology based Systems 1st Edition John Davies https://ebookgate.com/product/semantic-web-technologies-trends-and- research-in-ontology-based-systems-1st-edition-john-davies/ ebookgate.com Computational Intelligence for Movement Sciences Neural Networks Support Vector Machines and Other Emerging Technologies Rezaul Begg https://ebookgate.com/product/computational-intelligence-for-movement- sciences-neural-networks-support-vector-machines-and-other-emerging- technologies-rezaul-begg/ ebookgate.com
- 3. Decision Support Systems for Business Intelligence 2nd Edition Vicki L. Sauter https://ebookgate.com/product/decision-support-systems-for-business- intelligence-2nd-edition-vicki-l-sauter/ ebookgate.com Semantic Web Technologies and E Business Toward the Integrated Virtual Organization and Business Process Automation 1st Edition A. F. Salam https://ebookgate.com/product/semantic-web-technologies-and-e- business-toward-the-integrated-virtual-organization-and-business- process-automation-1st-edition-a-f-salam/ ebookgate.com Formal Ontologies Meet Industry 1st Edition R. Ferrario https://ebookgate.com/product/formal-ontologies-meet-industry-1st- edition-r-ferrario/ ebookgate.com On Flat Ontologies and Law First Edition Micha■ Dudek https://ebookgate.com/product/on-flat-ontologies-and-law-first- edition-michal-dudek/ ebookgate.com Semantic Service Integration for Smart Grids 1st Edition S. Rohjans https://ebookgate.com/product/semantic-service-integration-for-smart- grids-1st-edition-s-rohjans/ ebookgate.com
- 5. ONTOLOGIES AND SEMANTIC TECHNOLOGIES FOR INTELLIGENCE Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 6. Frontiers in Artificial Intelligence and Applications FAIA covers all aspects of theoretical and applied artificial intelligence research in the form of monographs, doctoral dissertations, textbooks, handbooks and proceedings volumes. The FAIA series contains several sub-series, including “Information Modelling and Knowledge Bases” and “Knowledge-Based Intelligent Engineering Systems”. It also includes the biennial ECAI, the European Conference on Artificial Intelligence, proceedings volumes, and other ECCAI – the European Coordinating Committee on Artificial Intelligence – sponsored publications. An editorial panel of internationally well-known scholars is appointed to provide a high quality selection. Series Editors: J. Breuker, N. Guarino, J.N. Kok, J. Liu, R. López de Mántaras, R. Mizoguchi, M. Musen, S.K. Pal and N. Zhong Volume 213 Recently published in this series Vol. 212. A. Respício et al. (Eds.), Bridging the Socio-Technical Gap in Decision Support Systems – Challenges for the Next Decade Vol. 211. J.I. da Silva Filho, G. Lambert-Torres and J.M. Abe, Uncertainty Treatment Using Paraconsistent Logic – Introducing Paraconsistent Artificial Neural Networks Vol. 210. O. Kutz et al. (Eds.), Modular Ontologies – Proceedings of the Fourth International Workshop (WoMO 2010) Vol. 209. A. Galton and R. Mizoguchi (Eds.), Formal Ontology in Information Systems – Proceedings of the Sixth International Conference (FOIS 2010) Vol. 208. G.L. Pozzato, Conditional and Preferential Logics: Proof Methods and Theorem Proving Vol. 207. A. Bifet, Adaptive Stream Mining: Pattern Learning and Mining from Evolving Data Streams Vol. 206. T. Welzer Družovec et al. (Eds.), Information Modelling and Knowledge Bases XXI Vol. 205. G. Governatori (Ed.), Legal Knowledge and Information Systems – JURIX 2009: The Twenty-Second Annual Conference Vol. 204. B. Apolloni, S. Bassis and C.F. Morabito (Eds.), Neural Nets WIRN09 – Proceedings of the 19th Italian Workshop on Neural Nets Vol. 203. M. Džbor, Design Problems, Frames and Innovative Solutions Vol. 202. S. Sandri, M. Sànchez-Marrè and U. Cortés (Eds.), Artificial Intelligence Research and Development – Proceedings of the 12th International Conference of the Catalan Association for Artificial Intelligence ISSN 0922-6389 (print) ISSN 1879-8314 (online) Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 7. Ontologies and Semantic Technologies for Intelligence Edited by Leo Obrst The MITRE Corporation, McLean, Virginia, USA Terry Janssen Lockheed Martin Corporation, Herndon, Virginia, USA and Werner Ceusters The State University of New York at Buffalo, Buffalo, New York, USA Amsterdam • Berlin • Tokyo • Washington, DC Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 8. © 2010 The authors and IOS Press. All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without prior written permission from the publisher. ISBN 978-1-60750-580-8 (print) ISBN 978-1-60750-581-5 (online) Library of Congress Control Number: 2010930895 Publisher IOS Press BV Nieuwe Hemweg 6B 1013 BG Amsterdam Netherlands fax: +31 20 687 0019 e-mail: order@iospress.nl Distributor in the USA and Canada IOS Press, Inc. 4502 Rachael Manor Drive Fairfax, VA 22032 USA fax: +1 703 323 3668 e-mail: iosbooks@iospress.com LEGAL NOTICE The publisher is not responsible for the use which might be made of the following information. PRINTED IN THE NETHERLANDS Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 9. Preface This book had its origin at the Second International Ontology for the Intelligence Community (OIC) Conference, which was held on November 28–29, 2007, in Colum- bia, MD, USA. At that time, a volume was proposed by the editors that would feature chapters by selected authors from the conference, who could extend their OIC papers or write on related topics that fit the guidelines the editors established for this book. In addition, other authors were invited to submit chapters. This book represents a partial technology roadmap for government information technology decision makers for information integration and sharing, and situational awareness (improved analysis support) in the use of ontologies, and semantic technolo- gies for intelligence. The general themes of both the OIC conferences and this book focus on intelli- gence community needs and the applications of ontologies and semantic technologies to assist those needs. Among the very many IC needs are the following: • To increase the ability to meaningfully share information, within and among communities, across humans and machines • To off-load some human cognitive functions and enable machines to assume these. By using ontologies and semantic technologies, machines come up to the human conceptual level, rather than humans having to go down to the ma- chine level, which latter tack has largely defined information technology since its orgins up to this point. • To increase the ability to automate some aspects of intelligence analysis, as for example, by supporting evidence-based reasoning, deductive (what logi- cally follows, given the knowledge) and abductive (what is the best explana- tion, given the evidence) queries • To provide assistance on probability of Hypothesis given the Evidence P(H|E), hypothesis generation, and analysis of competing hypotheses by using com- plex knowledge and logical mechanisms, and evaluating the consequences or ramifications of hypotheses • To increase the capability to semantically integrate data from all intelligence disciplines • To provide analytical tools that exploit the availability of semantically inte- grated information and knowledge • To assist in semantic disambiguation, reference, co-reference/correlation of entities, relations, and events o Disambiguation: To determine the appropriate meaning for the given con- text o Reference: To determine the actual entity in the world that the data refers to Ontologies and Semantic Technologies for Intelligence L. Obrst et al. (Eds.) IOS Press, 2010 © 2010 The authors and IOS Press. All rights reserved. v Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 10. o Co-reference/correlation: To determine whether two entities are actually the same entity, and the properties and events those entities respectively possess and participate in • To support reasoning over geospatial, temporal, and other data to infer addi- tional information about the real world. The target audience of this book is the US and other intelligence communities (IC), including law enforcement and homeland security communities, along with other tech- nical and budgetary decision makers and technologists working in intelligence. These technologists include ontologists and ontology developers, computer scientists, soft- ware engineers, and intelligence analysts who have a strong interest in semantic tech- nologies and their applications. This book would not have been possible without the assistance, dedication, and pa- tience of many generous individuals. We thank the IOS Press publisher and its dedi- cated representive Maarten Fröhlich for tolerance of delays in the editing of this book, while also providing constant and continuing encouragement. We thank the very many anonymous reviewers who helped improve the contributions of the authors by offering sound feedback and critical comments on multiple iterations of chapters. We thank the past organizers of the OIC conferences, for valuable suggestions and help on many issues, including in particular Barry Smith, Kathryn Blackmond Laskey, Duminda Wi- jesekera, and Paulo Cesar G. da Costa. We also thank Kevin Lynch and David Roberts, who provided governmental support for the OIC conferences and also feedback to the authors and editors on the impact of these technologies on the intelligence community, thereby serving to provide a pragmatic perspective to constrain the potential techno- logical exuberance. Of course the editors also thank their friends and families, who have countenanced aggravation, missed social opportunities, and personal inattention, to enable the writing and editing of this volume. Finally, we underscore that the views expressed in this book are those of the au- thors alone and do not reflect the official policy or position of The MITRE Corporation, the Lockheed-Martin Corporation, or any other company or individual, nor that of any particular intelligence community, agency, organization, or government. Leo Obrst Terry Janssen Werner Ceusters vi Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 11. Contents Preface v Leo Obrst, Terry Janssen and Werner Ceusters Chapter 1. Introduction: Ontologies, Semantic Technologies, and Intelligence 1 Terry Janssen, Leo Obrst and Werner Ceusters Chapter 2. How to Track Absolutely Everything 13 Werner Ceusters and Shahid Manzoor Chapter 3. Uses of Ontologies in Open Source Blog Mining 37 Brian Ulicny, Mieczyslaw M. Kokar and Christopher J. Matheus Chapter 4. A Multi-INT Semantic Reasoning Framework for Intelligence Analysis Support 57 Terry Janssen, Herbert Basik, Mike Dean and Barry Smith Chapter 5. Ontologies for Rapid Integration of Heterogeneous Data for Command, Control, & Intelligence 71 Leo Obrst, Suzette Stoutenburg, Dru McCandless, Deborah Nichols, Paul Franklin, Mike Prausa and Richard Sward Chapter 6. Ontology-Driven Imagery Analysis 91 Troy Self, Dave Kolas and Mike Dean Chapter 7. Provability-Based Semantic Interoperability for Information Sharing and Joint Reasoning 109 Andrew Shilliday, Joshua Taylor, Micah Clark and Selmer Bringsjord Chapter 8. The Use of Ontologies to Support Intelligence Analysis 129 Richard Lee Chapter 9. Probabilistic Ontologies for Multi-INT Fusion 147 Kathryn Blackmond Laskey, Paulo C.G. Costa and Terry Janssen Chapter 10. Design Principles for Ontological Support of Bayesian Evidence Management 163 Michael N. Huhns, Marco G. Valtorta and Jingsong Wang Chapter 11. Geospatial Ontology Trade Study 179 James Ressler, Mike Dean and Dave Kolas Chapter 12. Ontologies, Semantic Technologies, and Intelligence: Looking Toward the Future 213 Leo Obrst, Werner Ceusters and Terry Janssen Subject Index 225 Author Index 227 vii Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 12. This page intentionally left blank Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 13. Chapter 1 Introduction: Ontologies, Semantic Technologies, and Intelligence Terry JANSSENa,1 , Leo OBRSTb , Werner CEUSTERSc a Lockheed Martin Corporation, USA b The MITRE Corporation, c State University of New York at Buffalo Abstract: In recent years ontologies and semantic technologies more generally have begun to be applied to assist the intelligence community, for information integration, information-sharing, decision-support, and in many other applications. This chapter introduces the topic of the book and provides background information concerning its rationale, historical perspective, a vision for the future, and briefly describes the chapters of the present volume. Keywords: Ontology, information-sharing, intelligence community, semantic technologies. 1. Why Ontologies: What do the Intelligence Community and Its Customers Actually Need? Probably the best way to address this question is look at what the Director of National Intelligence (DNI) said in 2008 about the near future of intelligence, in the report titled The DNI’s Vision 2015 [1]. We are unable to do justice to this insightful document in this Introduction so the reader is encouraged to read it in its entirety. Some important points are pulled from this report, not in perfect context, and are quoted here [1]:2 In this [adversarial, terrorism] environment [worldwide] the key to achieving strategic advantage is the ability to rapidly and accurately anticipate and adapt to complex challenges… [p. 6] By 2015 we will need integrated and collaborative capabilities that can anticipate and rapidly respond to a wide array of threats and risks… [p. 7] To succeed in this fast-paced, complex environment, the Intelligence Community must change significantly. For example, our counterintelligence activities face an array of new and traditional adversaries, yet we must operate within a protected information-sharing environment that challenges existing notions of security [and] of risk… [p.7] Analytic precision and accuracy will be merely the minimum requirements expected by our customers; our accuracy must be clear, transparent, objective and intellectually rich… they will expect instantaneous support ‘on demand’. 1 Corresponding Author: Terry Janssen, Lockheed Martin Corporation, IS&GS/GSS, Herndon, VA 20171, USA; E-mail: terry.janssen@lmco.com. 2 Terms in bold are in the original document. Ontologies and Semantic Technologies for Intelligence L. Obrst et al. (Eds.) IOS Press, 2010 © 2010 The authors and IOS Press. All rights reserved. doi:10.3233/978-1-60750-581-5-1 1 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 14. By 2015, a globally networked Intelligence Enterprise will be essential to meet the demands for greater forethought and improved strategic agility. [But is important to note that] the purpose of intelligence is not merely to determine truth, but to enable decision-makers to make better choices in dealing with forces outside of their control. Intelligence helps reduce the degree of uncertainty and risk when crucial choices are made. Our measure of success is simple: did our service result in a real, measurable advantage to our side… [p. 10] [We] also need to exploit commercial technologies to develop new ways of providing service. [p. 11] By 2015, the Intelligence Community will be expected to provide more details about more issues to customers. We anticipate different types of customers – with greater expectations – and new demands to change the basic engagement model by which we serve them… [p. 11] To engage customers effectively, we must use sophisticated techniques… to elicit ‘What do you want to accomplish?’ … [and intelligence collection and analysis] will become more of an relationship than an event… [p. 12] Our analytic products will increasingly resemble customized services with an emphasis on maximum utility rather than simple releasability. Under concepts such as effects-based analysis, we will engage customers with ‘What-if?’ considerations in addition to ‘What?’ conclusions. To do so, our analysts will leverage disparate data and analytic tools and services, working in mission-focused distributed analytic networks… [p. 12] To respond to the dynamic and complex threat environment of the 21st century, our operating model must emphasize mission integration – a networked knowledge sharing model that rapidly pulls together disperse and diverse expertise and resources against specific missions… [p. 13] [It will] require multiple integrated collection systems… [and a] fully integrated processing, exploitation and dissemination architecture… [p. 14] There will be more emphasis on multi-agency teams pursuing ‘multi-INT’ collection strategies… We envision a collection community capable of rapidly fielding technological innovations that contain needed information… Above all else is the demand that the information reach those who need it, when they need it, in a form that they can easily absorb. [p. 14] The analytic community will be expected to understand and make judgments on a broad spectrum of national security threats, support a more diverse customer set, and cope with unprecedented amounts and types of information. Information overload already presents a profound challenge… [and] the analytic community has no choice but to pursue major breakthroughs in capability. Applying the principle of Collaborative Analytics analysts will be freed to work in a fundamentally different way – in distributed networks focused on a common mission. [p. 15] Information overload will be averted through sophisticated data preparation and tools. In 2015, new information will be tagged so tools can trace our data across our holdings. Analysts will use such tools to mine the data, to test hypotheses, and to suggest correlations. [p. 15] By 2015 the focus should shift from information sharing (e.g., interoperable systems, information discovery and access) to knowledge sharing [using an automated approach to the extent possible, with the humans in the loop to understand and present it accurately, and end-users to make decisions and act upon that knowledge]. [p. 17] Although we will continue to rely on commercial best of breed technologies and best practices, the Intelligence Community will still need to research, development and field disruptive technologies to maintain a competitive advantage over our adversaries. We cannot evolve into the next generation ‘S curve’ incrementally; we need a revolutionary approach… [p. 18] Creating a culture of innovation will require greater focus on advanced concepts, technology and doctrine to enhance leadership, organization alignment, and resources. [p. 19] T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 2 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 15. 2. Ontologies: An Enabling Technology Meeting the vision put forward by the DNI requires a fundamental shift in perspective from data sharing to knowledge sharing. It takes no more than a moment’s reflection to realize that knowledge cannot be shared unless it can be represented and communicated. In other words, interoperation at the knowledge level means more than syntactic interoperability and the sharing of data represented in standardized formats. The receiving system must interpret the data in the manner intended by the sending system. This means either that the communicating systems use common vocabularies with agreed-upon meanings for the terms, or that the interchange be mediated by an appropriate translation capability. It means that semantic information must be explicitly represented in a form accessible to all parties to a communication, so that a shared understanding of the knowledge being transmitted can be assured. This is precisely the purpose served by formal ontology. Ontologies represent the types of entities that exist in a domain, the relationships in which they can participate, and the attributes of entities of different types. Publicly available formal ontologies provide the basis for semantic interoperability by providing standardized representations to define the semantics of knowledge being exchanged. It is the thesis of this book that semantic technology used to address problems in the intelligence community is one of the “disruptive technologies” needed to maintain our competitive edge. Application of current-generation semantic technology can provide an immediate benefit. In addition, the process of developing applications will inevitably reveal important issues for which new research is needed, thus spurring advances in technology that will result in further practical benefits. 3. A Resource: The National Center for Ontological Research (NCOR) Recognizing the need for institutionalized leadership in semantic technology, the National Center for Ontological Research (NCOR) [2] was founded in October 2005 as a partnership of groups and institutions engaging in ontological research in the United States, with the State University of New York at Buffalo and Stanford as principal administrative sites. NCOR was established to serve as a vehicle to coordinate and to enhance ontological research activities, with a special focus on the establishment of tools and measures for quality assurance of ontologies, on training in and dissemination of good practices in the ontology field, and on the creation of strategies to advance the creation of federations of principles-based ontologies which work well together within the hub-and-spokes framework of the sort currently being advanced within the US Government’s Universal Core (UCore) and Command and Control (C2) Common Core (C2Core) initiatives [3, 4]. In 2008 NCOR was contracted by the US Army Net-Centric Data Strategy Center of Excellence to create a series of ontologies for use in the biometrics and C2 (command and control) domains, and also to create a standard Common Upper Ontology based on BFO and DOLCE, for the representation of entities in real-world domains. In 2009 NCOR worked with MITRE to develop UCore-SL, a Semantic Layer for UCore 2.0, an XML-based vocabulary resource designed to support data sharing sharing between agencies within the Department of Defense, Department of Homeland Security, Department of Justice and Department of Energy. T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 3 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 16. UCore and C2Core (and possibly other “common cores” in the future, depending on need) are vocabularies that exist in the Community of Interest (COI) paradigm within the federal government [5, 6]. In the COI paradigm, a modular architecture, as depicted in Figure 1, acts as structure for the emerging range of vocabularies. UCore spans all vocabularies and in particular immediately spans all common core vocabularies such as C2Core. The common core vocabularies in turn span all appropriate, COI vocabularies. COI vocabularies involve narrower domains and can be hierarchically structured, as shown in the figure. A COI comes into existence when two communities ascertain that they need to share information. The two communities engage in a discussion of the kinds of data they have and wish to share, and the vocabularies they use to refer to that data. Then they evolve an agreed upon vocabulary for the data they wish to share, thus developing a specific COI vocabulary. Figure 1. Universal Core, Common Cores, and COI Vocabularies In addition some COIs will develop semantic models of their vocabularies, i.e., ontologies. Others will develop structural models in XML Schema. An example of an ontology developed for a vocabulary is that of UCore-Semantic Layer (UCore-SL), an ontology that provides a semantics for UCore [7, 8]. At this time, UCore-SL is not officially part of UCore, but is a module under UCore Affiliates. Also, see [9, 10] for an early advocacy for common Command and Control semantics. When Figure 1 is compared with Figure 2 [11], a typical rendition of the layers of ontologies, one notices that there is somewhat of a correspondence between the layered vocabularies and the layered ontologies. However, in actuality, UCore addresses objects that typically would reside in a mid-level or lower upper level ontology, i.e., person, organization, facility, location, etc., and their properties. It is expected that UCore-SL would itself be embedded under an upper (sometimes called “foundational”) ontology such as Basic Formal Ontology (BFO) [12, 13], which is discussed in more detail below. T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 4 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 17. Figure 2. Ontology Architecture and Layers 4. A Resource: The Center of Excellence in Command, Control, Communications, Computing and Intelligence (C4I Center) C4I systems are essential to our national security. Recognizing the need for a strong intellectual base for C4I, and the need for comprehensive educational programs in C4I, George Mason University established the C4I Center in 1989 as the nation's first and only civilian university-based entity offering a comprehensive academic and research program in military applications of information technology. The Center performs research and supports educational programs in a wide variety of C4I areas. A central element in the C4I Center’s research is the formal representation of knowledge about the military and intelligence domains in both interoperable and machine processable forms. The Center is actively engaged in research to apply probabilistic ontologies to predictive naval situation awareness, and is developing an open source probabilistic ontology editor and reasoner. Another strong area of research is Battle Management Language (BML), a formalism to support reasoning about military doctrine and Command and Control (C2) processes, explicitly representing military task-based operations [14]. 5. An Ontology Success Story: The Peculiar History of the Gene Ontology The most conspicuous successes in ontology technology thus far have been in the biomedical field, and they result especially from the fact that the Gene Ontology (GO) [15] has been so widely used as a resource for the integration of data in the domains of molecular biology, bio-chemistry, functional genomics, proteomics, and related fields now of increasing relevance to clinical research and treatment. It is noteworthy that, in the period 2000-2007 there has occurred a 17-fold increase in use of the term T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 5 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 18. ‘ontology’ in the abstracts collected in the standard PubMed database of medical literature, yet almost all of this increase is associated with references to the Gene Ontology. There exist over 11 million annotations relating gene products described in major databases of molecular biology to terms in the GO. These annotations create linkages between genes and proteins to specific types of biological phenomena. Data related to some 180,000 genes have been manually annotated in this way, and the GO is hereby making the results of divergent kinds of life science research comparable and integratable. The GO is a founding partner of NCOR, and NCOR has played an important role in creating the OBO (Open Biomedical Ontologies) Foundry, a federation of the GO and its sister ontologies used in biomedical research [16]. The OBO Foundry is now serving as platform for the testing of NCOR strategies for quality assurance and ontology integration. The goal of the Foundry initiative is to create the conditions under which the data generated through biomedical research and clinical care will form a growing pool, to which algorithmic techniques can be applied in ways which serve the formulation and testing of clinical hypotheses at all levels. Efforts at ontology building are still standardly conceived in pragmatic terms, as projects motivated by the need to solve problems internal to the information technology needs of specific groups or organizations. The Foundry, in contrast, reflects a view of ontologies which sees them as lying outside the realm of software artifacts created to address specific local needs and sees them rather as part and parcel of the scientific enterprise [17]. Ontologies are from this perspective resources developed for the long term, freely available for use and subject to constant criticism and update. 6. An Example Initiative: The Basic Formal Ontology The hub, in the OBO Foundry and in a series of related initiatives, is Basic Formal Ontology (BFO), a top-level ontology building on lessons learned from the develop- ment of ontologies by logicians and philosophers over more than two millennia. BFO was developed initially to support the work of experimental scientists but is now increasingly gaining acceptance as a top-level ontology standard for general use. When ontologies are developed, like database schemas, simply to address local purposes, this will not only bring limited advantages in data integration but is indeed likely to intensify the very problems of forking which ontologies were designed to counteract. BFO is designed to serve as a top-level ontology standard that will constrain the developers of domain ontologies in such a way as to work against these effects. It is designed to be a very small, a true top-level ontology. This means that, in contrast to the foundational ontologies DOLCE [18] and SUMO [19], with which it otherwise has many features in common, BFO contains no terms which would properly belong within the domains of those spoke ontologies which extend it. Use of BFO in a hub-and- spokes framework thereby establishes a clear division of expertise. It provides both a simple common starting point for scientists in creating their ontologies, and also a common set of guidelines for ensuring that ontologies are thereafter developed and maintained in tandem with each other. BFO and its associated guidelines for ontology development have been refined on the basis of experience in application in the context of the OBO Foundry, and they are now increasingly being used also outside the scientific domain. T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 6 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 19. 7. Training and Dissemination of Ontology Technology In addressing its training and dissemination roles, NCOR has organized a series of ontology tutorials and training workshops, working closely in this also with the National Center for Biomedical Ontology. Jointly with the Ontolog Forum, the National Institute for Standards and Technology (NIST), and the newly founded International Association for Ontology and its Applications (IAOA) [20], NCOR organizes the annual Ontology Summit, held at NIST in Gaithersburg, MD since 2005. The theme for Ontology Summit 2010 is “Creating the Ontologists of the Future,” with a focus on certification individual ontologists and accreditation of institutions for courses of study for ontologists [21]. Jointly with JCOR, the Japanese Center for Ontological Research, NCOR organizes the InterOntology (Interdisciplinary Ontology) conference series held in Tokyo since 2006. Most importantly for our purposes, here, however, is the Ontology for the Intelligence Community (OIC) conference, discussed below, which has become a vibrant yearly forum for exchange of ideas on the role of semantic technology for problems of interest to the intelligence community. 8. Ontology for the Intelligence Community (OIC): A Forum for Knowledge Interchange The first two OIC conferences were organized by NCOR Director Barry Smith, and were held in Columbia, MD, in 2006 and 2007 [22] under the heading “Towards Effective Exploitation and Integration of Intelligence Resources.” The third and fourth OIC conferences were held in 2008 and 2009, respectively, at George Mason University under the auspices of the George Mason University C4I Center [23, 24]. The OIC series was established to support the work of those who are using ontologies to develop approaches to the analysis of intelligence that will enable greater flexibility, precision, timeliness and automation of analysis and thereby maximize valuable human resources in responding to fast-evolving threats. The OIC meetings have brought together researchers and intelligence analysts from major agencies, universities and other bodies involved in intelligence activities throughout the world. It provides an important venue for exchange of ideas and sharing of insights on the role and effective use of ontologies to problems in the intelligence domain. Speakers at the first two OIC conferences included Werner Ceusters, Director of the Ontology Research Group in the New York State Center of Excellence in Bioinformatics and Life Sciences in Buffalo, who described how BFO-based ontologies are being used to support the integration of instance data to enable tracking objects of all kinds in computer representations; representatives from the FBI, NSA, CIA, other USA agencies, the UK Defence Science and Technology Laboratory, and pivotal technologists involved in applying ontology and semantic technologies to intelligence needs; and Todd Hughes of the USA Defense Advanced Research Projects Agency (DARPA). OIC 2008 invited speakers were Deborah McGuinness of Rensselaer Polytechic Institute, one of the authors of the OWL Web Ontology Language; Michael Gruninger of the University of Toronto, ontology researcher and active participant in the ISO Common Logic standardization effort; and Leo Obrst, lead of the Information Semantics Group at MITRE Corporation. The invited speakers at OIC 2009 were Chris T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 7 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 20. Welty of IBM Watson Research Center, Doug Lenat of Cycorp, and a tutorial was presented by Leo Obrst. 9. Semantic Interoperability for DoD and IC Systems: BML and Beyond Modern military operations are increasingly becoming a joint endeavor, and it is hard to imagine any situation in which a single service would operate alone. Instead, most real world scenarios involve more than one country with multiple services, creating an urgent need for international standards to support coalition interoperability. BML is designed as an unambiguous language used to: 1) command and control forces and equipment conducting military operations; and 2) to provide situational awareness and a shared common operational picture. It can be seen as a standard representation of a digitized commander's intent to be used for real troops, for simulated troops, and for future robotic forces. BML is particularly relevant in a network centric environment for enabling mutual understanding. A BML development focus has always been conveying doctrinal knowledge among military forces with diverse C2 processes. Thus, the advantages are clear of evolving BML into a full ontology on military operations based on a strong formalism for reasoning about tasks and actions. Such a C2 ontology would be a useful complement to an intelligence community ontology. Achieving semantic interoperability among DoD and IC systems is essential, but one must recall that these systems will be always operating under the “fog of war”, where incomplete data is the rule and uncertainty is ubiquitous. Unfortunately, current ontology technologies provide no support for representing and reasoning in a principled way with uncertainty and incomplete data. PR-OWL, a Bayesian first-order logic extension to the ontology language OWL, was developed at the GMU C4I Center, and is an example of current efforts to build probabilistic-aware ontologies [25]. These and other similar efforts can be seen promising enabling technologies for the vision set forth in the DNI’s Vision 2015. 10. The Contributions There are a number of themes in the chapters of this book. Many of these themes span multiple chapters, and many chapters have multiple themes. For example, many of these chapters focus on supporting intelligence analysts using semantic technologies and develop proofs of concept, especially the earlier chapters. The Ceusters and Manzoor chapter, the Ulicny, Kokar, and Matheus chapter, Obrst et al, Lee, Shalliday et al, Self et al, Ressler et al primarily address this theme. Another theme, however, is that of dealing with uncertainty in ontology-based technologies, and hence addressing the interaction between ontology and epistemology. Chapters that focus on this theme include the Janssen et al, Laskey et al, and Huhns et al chapters. Finally, the first chapter (the current chapter) and the book’s final chapter focus primarily on the impact of ontologies and semantic technologies on intelligence collection and analysis – the former attending to past and current efforts, the latter addressing issues about and potential impacts on the future. The final chapter is that by Obrst, Ceusters, and Janssen. T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 8 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 21. The individual chapters address significant issues for intelligence by focusing on particular aspects of ontologies and semantic technologies. In Chapter 2, “How to Track Absolutely Everything” Ceusters and Manzoor develop a referent tracking paradigm that tracks ontology-based unique real entities and events, but also information and data elements used in information systems to describe these. By using globally unique identifiers, knowledge and assertions about the real world, beliefs and cognitive representations based on observations of collectors, analysts, etc., and metadata including provenance, can be tracked over time, through many changes. Such a referent tracking facility would greatly extend the current capabilities of data and metadata repositories, for example. In Chapter 3, the “Uses of Ontologies in Open Source Blog Mining” (Ulicny, Kokara, Matheus), the authors make the case for using ontologies to mine blog entries, which are very dynamic and difficult to automatically interpret, aggregate, and report on. As with other structured and unstructured online data, an analyst cannot read everything, so if semantic tools can enable him/her to classifiy (i.e., bin in finer granular bins with certain topics and properties of interest) the content, the machine can better assist the analyst at finding truly relevant information. The need to find information in blogs is great enough that a number of blog-specific search engines have arisen in recent years, in including Technorati, BlogPulse. In Chapter 4, “A Multi-INT Semantic Reasoning Framework for Intelligence Analysis Support” (Janssen, Basik, Dean, Smith), a possible approach to the information overload that afflicts intelligence analysts is to augment human capabilities of winnowing, interpreting, and integrating data by enlisting machines that use ontologies and other semantic technologies. Software can thereby perform lower-level knowledge functions that are comparable to what a human would perform, draw reasonable human-like inferences over that knowledge, and then present the interesting subsets of knowledge that analysts would be most interested in, and establish linkages among those knowledge components. Because the knowledge needed for many intelligence problems is a fusion of information from many intelligence disciplines, the authors developed a multi-INT ontology and framework using the HighFleet (formerly known as Ontology Works, Inc.) knowledge server. In Chapter 5, “Ontologies for Rapid Integration of Heterogeneous Data for Command, Control, & Intelligence” (Obrst, Stoutenburg, McCandless, Nichols, Franklin, Prausa, Sward), the authors present a program that uses ontologies expressed in the Semantic Web Ontology Language OWL and rules expressed in the Semantic Web Rule Language (SWRL) to provide efficient runtime reasoning for situational awareness and course of action assistance. The authors’ prototype is focused on convoy movement in a theater of operation, where the convoy has a primary path (and possibly other paths) from its origin to its destination, and potentially encounters many hostile or unknown theater objects, which may impact the convoy’s mission. The ontologies focus on theater objects and intelligence information, while the rules focus on detecting theater objects, based on incoming intelligence, determining their impact on the convoy, and then alerting the convoy to their presence, and making some recommendations for their avoidance (for example, by changing course, or assuming a defensive posture if the theater object cannot be outrun). The ontologies and rules are transformed and then compiled into a logic programming engine. Using an enterprise service bus to link multiple instances of the logic programming reasoner and providing a terrain-oriented visualization based on Google Earth, the tool enables the convoy commander to employ high-level machine assistance for recognizing potentially hazardous situations and T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 9 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 22. thereby potentially improve decision making. Finally, the chapter shows how the initial set of ontologies was extended for other applications, such as for space objects and events and for unmanned aerial vehicle (UAV) flight deconfliction. Chapter 6, “Ontology-Driven Imagery Analysis” (Self, Kolas, Dean) addresses the improvement of imagery analysis by annotating images with terms from ontologies. A software imagery analysis environment is described which provides a way to record structured annotations to images based on their semantics, and so enabling much richer search and more powerful exploitation of imagery. For example, using semantic annotations of images may enable images to be combined more meaningfully, by determining if sets of observations are related. How are different images related over time? What has changed from a sequence of images over time? Are there correspondences or differences between related geospatial regions over time? How can one efficiently query semantically annotated image repositories? In Chapter 7, “Provability-Based Semantic Interoperability for Information Sharing and Joint Reasoning” (Shilliday, Taylor, Clark, Bringsjord), the authors describe their system for provability-based semantic interoperability that encodes a translation graph for the ontologies to be compared (including the same ontology as modified into different versions over time), using a many-sorted logic. They also situate their approach in the wider spectrum of approaches for addressing semantic interoperability, namely via the development of schema (ontology) mappings and schema (ontology) morphisms. By first creating signatures of the ontologies in a many-sorted logic, they are able to create a translation graph that visually depicts the incremental construction and interrelation of ontology signatures, showing the transformations that that map one ontology signature into another or into a different version of its prior self (as for versioning an ontology). The translation graph is therefore a directed graph with vertices of signatures and edges that represent the relationships between signatures. Finally, the authors provide an example of their framework in action, in the UAV domain, where multiple data-source ontologies need to be integrated. In Chapter 8, “The Use of Ontologies to Support Intelligence Analysis”, Richard Lee describes the Metadata Extraction and Tagging Service (METS) effort to support intelligence analysis by providing ontologies, i.e., semantic models, rather than simply XML structural models, for tagging data of interest to analysts, including entities obtained from information extraction over datasets. A multi-INT data fusion experiment is described, which highlights the limitations of XML-based, i.e., syntactic approaches. In Chapter 9, “Probabilistic Ontologies for Multi-INT Fusion” (Laskey, Costa, Janssen), the authors focus on multi-INT fusion of heterogeneous information sources using semantic resources such as ontologies, but primarily those extended with probabilities expressed in the Probabilistic OWL (PR-OWL) formalism. PR-OWL extends the Web Ontology Language OWL with probabilistic support from Bayesian semantics, so that complex patterns of evidential relationships among uncertain hypotheses can be represented and reasoned over by machine. An early version of a reasoning engine, UnBBayes-MEBN (where MEBN stands for Multi-Entity Bayesian Networks), is described, which support PR-OWL querying and inferencing. Chapter 10, “Design Principles for Ontological Support of Bayesian Evidence Management” (Huhns, Valtorta, Wang) takes the approach that indeed Bayesian semantics should be wedded to ontologies for management of evidence. Ontologies provide knowledge about the domain, events, and causality, and then Bayesian reasoning provides evidential reasoning in their Magellan system using fragments of T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 10 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 23. situations, which can be combined to provide sets of possible situations that are consistent with the evidence. The combination, according to the authors, provide analysts with a way to gauge competing hypotheses and reduce the uncertainty of their potential outcomes. Magellan uses XMLBIF (eXtensible Markup Language Bayesian Interchange Format), along with OWL, RDF, and the SPARQL query language. Chapter 11, “Geospatial Ontology Trade Study” (Ressler, Dean, Kolas) is focused on the ontologies and other semantic standards that are valuable to geospatial analysis, of particular interest to those involved with the GEOINT discipline, but also others who require geospatial reasoning. Geospatial semantics ranges over representation of geometry (i.e., points, lines, spaces, spheres, etc., involved in addressing regions), geopolitics (i.e., borders, locations defined politically), temporal notions (how geospatial notions change over time), and geographical and topographical knowledge. As part of their trade study, the authors provide a matrixed view of the semantic In the concluding chapter, Chapter 12: Ontologies, Semantic Technologies, and technologies for addressing intelligence analysis and collection. A discussion is presented on the use cases for the applications of these technologies vs. their complexity, as gauged by the required expressiveness of the semantic models needed to provide those applications. Cost must be addressed too, and measured against potential benefits, as some emerging ontology cost models intend to provide. In addition, emerging standards and technologies are discussed, from the SPARQL query language, triple stores (that are repositories of OWL/RDF instance data, structured in graphs), and rules and rule languages such as the Rule Interchange Format (RIF) – all of which are supported by rapidly emerging tools. A prospective lesson is given for intelligence that draws on the experience of using realist ontologies in biomedicine and healthcare, in the hope that some notion of the value of ontologies and semantic technologies can be indicated for intelligence collection and analysis. A distinction is then made between ontology (the ways things are) and epistemology (the ways things are believed to be or that we have current evidence for). Both technical disciplines and their tools are crucial for intelligence, and provide complementary value. A human being has only one birth date (ontology), but which of several ascribed to a particular person is correct (epistemology)? Finally, the authors express optimism about the emerging convergence of intelligence analysis and semantic technologies, and the potential value of that convergence for intelligence. REFERENCES [1] Vision 2015: A Globally Networked and Integrated Intelligence Enterprise. Office of the Director of National Intelligence (ODNI), July 22, 2008. http://www.dni.gov/reports/Vision_2015.pdf. [2] National Center for Ontological Research (NCOR). http://ncor.us/. [3] Universal Core (UCore). https://ucore.gov. [requires authorized login] [4] Command and Control Common Core (C2Core). https://www.us.army.mil/suite/page/473883. [requires authorized login] [5] Communities of Interest (COI) – Home, Assistant Secretary of Defense (Networks and Information Integration) Department of Defense Chief Information Officer (ASD(NII)/DoD CIO). http://www.defenselink.mil/cio-nii/sites/coi/coi.shtml.htm. described is that of the future and the potential promise of ontologies and semantic standards that are possibly most applicable and important to geospatial analysis. Intelligence Looking Toward the Future” (Obrst, Ceusters, Janssen), the view : “ T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 11 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 24. [6] DoD Directive 8320.2. Data Sharing in a Net-Centric Department of Defense. December 2, 2004, Certified Current as of April 23, 2007. http://www.defenselink.mil/cio- nii/sites/coi/Governance/832002p.pdf. [7] Smith, Barry; Lowell Vizenor; and James Schoening. 2009. Universal Core Semantic Layer. In Proceedings of the Ontologies for the Intelligence Community (OIC) conference, October 20-22, 2009, George Mason University, Fairfax, VA. [8] Smith, Barry. 2009. Universal Core Semantic Layer (UCore SL): An Ontology-Based Supporting Layer for UCore 2.0, UCore Conference, MITRE, McLean, VA. September 23, 2009. [9] Chaum, Erik; Richard Lee. Command and Control Common Semantic Core Required to Enable Net- centric Operations. AFCEA-George Mason University Symposium "Critical Issues in C4I" May 20-21, 2008. [10] Winters, Leslie and Andreas Tolk. 2009. C2 Domain Ontology within Our Lifetime. In Proceedings of the 14th International Command and Control Research and Technology Symposium (ICCRTS), Jun 15- 17, 2009, Washington, DC. [11] Semy, S.; Pulvermacher, M.; L. Obrst. 2005. Toward the Use of an Upper Ontology for U.S. Government and U.S. Military Domains: An Evaluation. MITRE Technical Report, MTR 04B0000063,November 2005. http://www.mitre.org/work/tech_papers/tech_papers_05/04_1175/index.html. [12]. Basic Formal Ontology (BFO). http://www.ifomis.org/bfo. [13] P Grenon, B Smith. 2004. SNAP and SPAN: Towards dynamic spatial ontology. Spatial Cognition and Computation: An Interdisciplinary Journal, Vol. 4, No. 1. (2004), pp. 69-104. [14] Battle Management Language (BML). http://c4i.gmu.edu/BML.php. [15] Gene Ontology. http://www.geneontology.org. [16] Open Biomedical Ontologies (OBO) Foundry. http://obofoundry.org. [17] Smith, Barry. “Ontology (Science)”, in C. Eschenbach and M. Gruninger (eds.), Formal Ontology in Information Systems. Proceedings of the Fifth International Conference (FOIS 2008), Amsterdam: IOS Press, 21-35. http://precedings.nature.com/documents/2027/version/2. [18] Descriptive Ontology for Linguistic and Cognitive Engineering . DOLCE. http://www.loa- cnr.it/DOLCE.html. [19] Suggested Upper Merged Ontology (SUMO): http://www.ontologyportal.org/. [20] International Association for Ontology and its Applications (IAOA). http://www.iaoa.org/. [21] Ontology Summit 2010: Creating the Ontologists of the Future. http://ontolog.cim3.net/cgi- bin/wiki.pl?OntologySummit2010. [22] Hornsby, Kathleen Stewart, ed. 2007. Proceedings of the Second International Ontology for the Intelligence Community Conference (OIC) 2007, November 28-29, 2007, Columbia, MD, USA. CEUR Workshop Proceedings, Volume-299. http://ftp.informatik.rwth-aachen.de/Publications/CEUR- WS/Vol-299/. [23] Laskey, Kathryn Blackmond; Duminda Wijesekera. 2008. Proceedings of the Third International Ontology for the Intelligence Community Conference. Fairfax, VA, USA, December 3-4, 2008. CEUR Workshop Proceedings Volume-440. http://sunsite.informatik.rwth-aachen.de/Publications/CEUR- WS/Vol-440/. [24] Costa, Paulo; Kathryn Laskey; Leo Obrst, eds. 2009. Proceedings of the 2009 International Conference on Ontologies for the Intelligence Community. Fairfax, VA, USA, October 21-22, 2009. CEUR Workshop Proceedings Volume-555. http://sunsite.informatik.rwth-aachen.de/Publications/CEUR- WS/Vol-555/. [25] Costa, Paulo; Kathryn Blackmond Laskey. 2006. PR-OWL: A Framework for Probabilistic Ontologies. Proceedings of the Fourth International Conference on Formal Ontology in Information Systems. November 2006. http://ite.gmu.edu/~klaskey/papers/FOIS2006_CostaLaskey.pdf. T. Janssen et al. / Introduction: Ontologies, Semantic Technologies, and Intelligence 12 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 25. Chapter 2 How to Track Absolutely Everything Werner CEUSTERS1 , Shahid MANZOOR Ontology Research Group, New York State Center of Excellence in Bioinformatics and Life Sciences, University at Buffalo, USA Abstract: The analysis of events prior to and during September 11 revealed that a smooth execution of the intelligence process is hampered by inadequate information sharing. This caused a rethinking of the intelligence process and a transition towards a ‘Globally Networked and Integrated Intelligence Enterprise’ with the goal that more detailed, tagged, and, therefore, traceable, information will reach those who need it, when they need it, and in a form that they can easily absorb. We present the referent tracking paradigm and its implementation in networks of referent tracking systems as an enabling technology to make this vision come true. Referent tracking uses a system of singular and globally unique identifiers to track not only entities and events in first-order reality, but also the data and information elements that are created to describe such entities and events in information systems. By doing so, it meets the requirements of the Nation’s Information Sharing Strategy. Keywords: referent tracking 1. Introduction Intelligence, as defined by the Central Intelligence Agency (CIA), is ‘the information our nation’s leaders need to keep our country safe’ [1]. This information is produced by the US Intelligence Community (IC), i.e. the departments and agencies cooperating to fulfil the goals of Executive Order 12333 which stipulates that ‘The United States intelligence effort shall provide the President and the National Security Council with the necessary information on which to base decisions concerning the conduct and development of foreign, defense and economic policy, and the protection of United States national interests from foreign security threats’ [2]. This is achieved through the performance of what is called the ‘intelligence process’ which consists of five steps: (1) the determination of the information requirements, (2) the collection of raw data, (3) the processing of the raw data into forms that are more usable for intelligence analysts or other consumers, (4) the integration, evaluation and analysis of the data in order to generate reports satisfying the requirements, and (5) the dissemination of the results to the appropriate level [3]. This last step, typically, leads to new information requirements which initiate a new cycle of the intelligence process. 1 Corresponding Author: Werner Ceusters, Ontology Research Group, New York State Center of Excellence in Bioinformatics and Life Sciences, 701 Ellicott street, Buffalo NY, 14203, USA; E-mail: ceusters@buffalo.edu Ontologies and Semantic Technologies for Intelligence L. Obrst et al. (Eds.) IOS Press, 2010 © 2010 The authors and IOS Press. All rights reserved. doi:10.3233/978-1-60750-581-5-13 13 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 26. 1.1. Challenges and barriers Ideally, the information that is finally disseminated is (1) reliable, thus corresponding faithfully to what is the case in reality, (2) complete, such that nothing what is essential or required for the consumer to make adequate decisions is missing, (3) relevant, such that decisions can be made efficiently, and (4) timely, guaranteeing that decisions can be made early enough for the resulting actions to have the desirable effect. Unfortunately, this ideal is very hard to achieve because of many barriers and challenges [4]. A large number of these challenges are brought about by the multiplicity of agencies, organizational levels within these agencies and information consumers that are involved. Although each step in the intelligence process comes with its own challenges, the multiplicity of involved actors affects primarily the information requirements assessment and the data-integration and analysis steps. So do the information requirements that a specific organizational level has to take into account not only consist of the external requirements put forward by the consumers to whom intelligence reports of a specific nature and content need to be delivered, but also of the internal requirements which determine what sorts of detailed information elements are required and accessible to provide high quality reports. The integration, evaluation and analysis step can be hampered by insufficient lower-level data (both quantitatively and qualitatively), wrong information, and lack of meaningful data linkage. The net effect is that the reliability, completeness and relevancy of the resulting conclusions suffer considerably. Although these three notions are intuitively straightforward, they can be defined in various ways and for each such way, objective quantification is hard, if possible at all. Furthermore, these notions are not entirely independent from each other. Reliability, for instance, relates to accuracy which itself relates to relevancy: the more a measurement is accurate, the more reliable it seems to be, yet, the relevancy of it might diminish depending on the objectives of the intelligence effort: whereas providing information on the duration of intercontinental flights in minutes to compare the performance of foreign carriers with that of national ones seems reliable, accurate and relevant, doing so in hours is hardly reliable, while in seconds for sure not relevant. Redundancy of information elements within a collection of information will not harm the completeness and relevancy of that collection as a whole, but for sure the relevancy of the redundant elements themselves. At the other hand, from a second order perspective, the presence of redundant information, if obtained from various independent sources, might be an indication for the reliability of the collection. 1.2. Intelligence and Security Informatics The analysis of events prior to and during September 11 revealed that a smooth execution of the intelligence process is hampered by inadequate information sharing [5]. Not only are there legal and cultural barriers to information sharing – the ‘need-to- know’ culture during the Cold War is now recognized to be a handicap in dealing with terrorism and other asymmetric threats [6] – it is also technically very difficult to integrate and combine data that are stored in different database systems running on different hardware platforms and operating systems [7]. Although the Office of Homeland Security, in 2002, identified information sharing across jurisdictional boundaries of intelligence and security agencies as one of the key foundations for W. Ceusters and S. Manzoor / How to Track Absolutely Everything 14 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 27. ensuring national security [8], the appropriate infrastructure is not yet there. This recognition led to the development of a new science: ‘Intelligence and Security Informatics’ (ISI) [9], which is commonly defined as ‘the study of the use and development of advanced information technologies, systems, algorithms, and databases for national- and homeland-security-related applications through an integrated technological, organizational, and policy-based approach’ [10]. ISI tries to overcome the barrier that data which reside in distinct data sources are organized in different schemas, and therefore are difficult to integrate. But still, once some sort of integration has been achieved, it remains often very hard to determine, for instance, whether two distinct pieces of information are about the same entity or which piece of information is correct when several pieces about the same entity can’t be true at the same time. As an example, a case study in a local police department revealed that more than half of the suspects had either a deceptive or an erroneous counterpart existing in the police system: 42% of the suspects had records alike due to various types of unintentional errors, while about 30% had used intentionally a false identity [11]. Deception is in the context of ISI a very hard problem indeed; it is not limited to providing false identities, but includes also ‘cognitive hacking’ which involves disinformation attacks on the mind of the end user of a networked computer system such as a computer connected to the Internet [12]. Identifying such attacks is crucial in an era in which the Intelligence Community seeks to make better use of Open Source Information (OSINT) [13]. 1.3. Vision 2015 To further advance the modernization of the information technology within the Intelligence Community, the Office of the Director of National Intelligence [14] published in February 2008 its ‘Information Sharing Strategy’ report [6], followed in July by the ‘Vision 2015’ document [15]. They key idea, first introduced in the National Intelligence Strategy [16], is the move towards a ‘Globally Networked and Integrated Intelligence Enterprise’ with the goal that more detailed, tagged, and, therefore, traceable, information will reach those who need it, when they need it, and in a form that they can easily absorb. Efforts in these directions are expected to create the ability to develop, digest, and manipulate vast and disparate data streams ‘about the world as it is today’ by means of tags that enable the use of tools that can ‘trace related data across our holdings, to mine the data, to test hypotheses and to suggest correlations’ in addition to ‘measuring performance’ [15]. The key characteristics of the new information sharing model are [6]: C1. ‘responsibility to provide’: sharing intelligence data while still addressing the need to protect privacy, civil liberties, and sources and methods; C2. enterprise-centric: providing services across agencies, partners, and international borders for multiple mission use; C3. mission-centric: able to adapt rapidly to changing needs and new partners; C4. information-centric: security built into the data and environment using tags; C5. attribute-based: access based on attributes that go beyond security classification (e.g. environmental, affiliation, mission focus, etc.); C6. data ‘stewardship’ (rather than data ‘ownership’), focusing on quality and reusability of data rather than, but not excluding, protection. W. Ceusters and S. Manzoor / How to Track Absolutely Everything 15 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 28. 1.4. Tagging, indeed, but what and how? Because ‘tagging’ seems to be an important part of the proposed solution to make this vision come true, the issues that we address here are (1) where the tags should come from, (2) what it is that should be tagged, and (3) according to what sort of logical schema data and tags should be organized in order for the data to track faithfully what is going on in the world. We argue, in response to each of the issues just mentioned, (1) that the tags should correspond to the terms (or codes) which are used as representations for universals and defined classes in realism-based ontologies, thus covering what is generic, (2) that what is tagged should not only be the data about first- order entities (persons, vehicle movements, parcels, disease outbreaks, …), but also how and by whom (and what) these data are generated and manipulated, and (3) that the data should be organized in a structure which mimics the structure of that part of reality that is described by the data and that is capable to reflect all sorts of changes that reality undergoes in the course of history. 2. Naive tagging Today, information is primarily maintained in information systems which consist of data repositories that contain data in either unstructured form (such as free text or digital multi-media objects) or structured form, the latter being such that numerical information is expressed by means of numbers, and non-numerical information by means of codes or terms associated with what is commonly called ‘concepts’, taken from different sorts of terminologies (such as vocabularies, nomenclatures, concept systems, and so forth) as they are offered in terminology servers. Since data in structured form are better suited to provide software agents with a deep understanding of what the data represent, considerable efforts are spent to turn unstructured data into structured data, at least partially. However, whether data are captured in structured form when entered, or rendered as such afterwards using text and image analytics software which add tags corresponding to concepts, current information systems exhibit at least two major shortcomings as far as concept-based tagging is concerned: (1) formal impreciseness about what is tagged, and (2) incompatibility of distinct tagging systems. 2.1. Missing the point(ers) Mainstream information systems do not offer a mechanism to unambiguously determine in each individual case what entity in reality a concept from a terminology server is used to relate to. As a consequence, information systems thus conceived work with instances of data, but algorithms working on such data have no clue what the data are about, i.e. about what specific entity in reality each specific data-element contains the information. If, for example, a driving license number is used in an information system, it is often not formally clear whether the number is used to denote the driving license of a person or that person itself. As a further example, if in an information system the gender of a person is stated to be ‘unknown’, then it is often not formally clear whether this means either (1) that the person does have a gender which is one of the scientifically known gender types W. Ceusters and S. Manzoor / How to Track Absolutely Everything 16 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 29. such as female, male, mosaic, etc., but that information of the precise gender of that person is not available in that information system, or (2) that the gender of that person is known to be of a type which scientifically has not yet been determined. Another example is that if at a certain time the gender of a specific person is registered in some information system as ‘male’, and at a later time as ‘female’, then there is, under existing data storage paradigms, no way to derive from this change whether the change in the information system reflects (1) a change in reality, for instance, because the person underwent transgender surgery, (2) a change in what became known about reality: the person’s gender might because of a congenital disorder not have been determinable at the time of birth, but only later after several investigations, or (3) that there was no change in reality or what we know about it, but that at the time of the first entry a simple mistake was made. One can even imagine a fourth possibility, namely that the meaning of the word ‘female’ would have been changed. The latter might seem to be too far fetched – in fact, this did never happen for the words ‘male’ and ‘female’ – but there are several examples in the past that come close. The title ‘Chief Executive Officer’, for instance, was introduced in Europe in the late eighties, replacing titles such as ‘Director General’ or ‘Managing Director’. A change in title, in those days, for sure did not entail a change in position or power of the person to whom the new title was attributed. These types of issues are insufficiently addressed in modern Semantic Web applications because they are not yet generally recognized: attempts to address them are sparse. 2.2. Missing semantics The most recent hype in information system networking is semantic interoperability. By ‘semantic interoperability’, it is meant the ability of two or more computer systems to exchange information and have the meaning of that information automatically interpreted by the receiving system accurately enough to produce useful results, as defined by the end users of both systems. Current attempts to achieve semantic interoperability rely on agreements about the meaning of so-called concepts stored in terminology-systems, such as nomenclatures, vocabularies, thesauri, or ontologies, the idea being that if all computer systems use the same terminology, they can understand each other perfectly. The reality is, however, that, rather than one such terminology being generally adopted, the number of terminology-systems with mutually incompatible definitions or non-resolvable overlap amongst concepts grows exponentially, thereby contributing more to the problem of semantic non- interoperability than solving it. Of course, ontologies developed for different purposes can only reasonably be expected to have partial overlap, but more efforts should be conducted to exploit overlap when resolvable. 3. Fundamentals of realism-based ontologies and data repositories In contrast to traditional terminology approaches, the realist orientation in terminology and ontology is based on the view that terms in terminologies are to be aligned not on concepts but rather on entities in reality [17]. Central to this view are three assumptions [18]. The first is that reality exists objectively in itself, i.e. independent of the perceptions or beliefs of cognitive beings. Thus not only do a wide variety of entities W. Ceusters and S. Manzoor / How to Track Absolutely Everything 17 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 30. exist in reality (human beings, terrorists, guns, attacks, countries, ...), but also how these entities relate to each other (that human beings are citizens of countries, that in most attacks guns are used, and so forth) is not a matter of agreements made by scientists or database modellers but rather of objective fact. The second assumption is that reality, including its structure, is accessible to us and can be discovered: it is scientific research that allows human beings to find out what entities exist and what relationships obtain between them. It is intelligence analysis that allows analysts to find out which specific human beings are terrorists. The third assumption is that an important aspect of the quality of an ontology or terminology is determined by the degree to which the structure according to which the terms are organized mimics the pre-existing structure of reality. In the context of information systems, it means that an important aspect of the quality of an information system is determined by the degree to which (1) its individual representational units correspond to entities in reality, and (2) the structure according to which these units are organized mimics the corresponding structure of reality. 3.1. Faithful representations The above assumptions form the basis for distinguishing between three levels of reality which have a role to play wherever ontologies are used as artifacts for annotation and tagging, and wherever automated or semi-automated reasoning is required to be able to deal with an overload of information, parts of which can be expected to be wrong. Ontologies and data repositories for the intelligence community are no exception to this. The three levels are [18]: • Level 1: the (first-order) reality ‘in the field’: the persons that are tracked, the events that are monitored, the users of the information system, and so forth; • Level 2: the beliefs and cognitive representations of this reality embodied in observations and interpretations on the part of observers, data collectors, analysts and others; • Level 3: the publicly accessible concretizations of such cognitive representations in representational artifacts of various sorts, of which ontologies, terminologies and data repositories are examples. Ontologies contain typically representations for what is generic, thus representing entities such as person, weapon, war, and so forth. Repositories cover what is specific, thus holding representations for entities such as President George W. Bush Jr., the gun that killed John F. Kennedy, The Gulf War, etc. In line with the theory of granular partitions [19] we argue that complex representations should be composed in modular fashion of sub-representations built out of representational units that are assumed to correspond to portions of reality (POR). Some characteristics of the units in a representation created for intelligence purposes are: • each such unit is assumed by the authors of the representation to be veridical, i.e. to conform to some relevant POR as conceived on the best understanding (which may, of course, rest on errors). Thus if in a data repository a representational unit standing proxy for a specific person is associated with the name ‘George Bush’, then, under the realist paradigm, we assume that a person with this name exists or has existed (that on the basis of the name only W. Ceusters and S. Manzoor / How to Track Absolutely Everything 18 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 31. it cannot be determined which specific person is meant, does not make the unit non-veridical); • several units may correspond to the same POR by presenting different though still veridical views or perspectives, for instance at different levels of granularity (one thing may be described both as being brown and as reflecting light of a certain wavelength, or one event as an event of administering and of consuming drugs); • what units are included in a representation depends on the purposes which the representation is designed to serve. 3.2. Keeping track of changes The real world is subject to constant change, and so also is our knowledge thereof. To keep track of these two sets of changes, any representation concerning a relationship between entities should be associated with at least the following pieces of information: (P1) an index for the time period during which the relationship obtains, (P2) an index for the time at which the representation is made, i.e. the time at which the relationship is (believed to be) known, (P3) an index for the time that piece of information is made available in the system, and (P4) an identifier standing proxy for the author of the representation. Keeping track of these various types of information makes it possible not only to track reality faithfully from an individual analyst or agency perspective, but also to preserve the knowledge about what was known by whom and at what time after information which was residing originally in distinct systems becomes merged. It also allows to assess whether information is disclosed in a timely fashion. Suppose, for instance, that at time t10 it is known by analyst A1 that suspect S was since t9 member of group G of possible terrorists, but that an entry to that effect in the information system of his agency is made available not earlier than at t11. Thus between t10 and t11, that information was not accessible. Furthermore, in reality, it might be that S was already member of G at t5. That information might have been known in another agency since t6, and made available at that time in their information system. When the information in the two systems becomes merged, for instance after the Vision 2015 situation becomes reality, it can still be assessed what was known at each point in time in each agency. 4. Fundamentals of Referent Tracking Referent Tracking (RT) is a paradigm for information management that is distinct from other approaches in that each data element has to point to a portion of reality in a number of predefined ways (Figure 1). It has been introduced in the context of Electronic Health Record keeping [20], but its applicability is wider than that, examples being digital rights management [21] and corporate memories [22]. By ‘portion of reality’ is meant any individual entity or configuration of entities standing in some relation to each other. By ‘entity’ is meant anything that exists or has existed in the past, whatever its nature. A ‘configuration’ is a portion of reality which is not an entity in its own right. Whereas a specific person, his or her activities, the social network he belongs to, the analyst examining information about that person, and that W. Ceusters and S. Manzoor / How to Track Absolutely Everything 19 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 32. examination itself are each individual entities, the configuration that the activities of this person are being monitored by an intelligence agency, or his or her being part of that social network, is not. Another example of a configuration is the being of an engine in a car. Both that car and that engine are entities, but the fact that that engine is in that car, is not. If that engine would not be in the car, but, for instance be placed by a mechanic outside the car for repair purposes, still the very same entities (the car and the engine) would be involved, but there would be another configuration. Within the RT paradigm, configurations are referred to by means of a data type called a ‘RT-tuple’, whereas entities are represented by means of a data type called ‘representation’. Both data types come in several forms depending on the nature of the portion of reality they carry information about (see section 6). RT, through its data types, allows also for the drawing of an explicit distinction made in Basic Formal Ontology (BFO) [23] between specific entities called ‘particulars’ from generic entities called ‘universals’. Particulars are specific and unique entities, unique in the sense that they each occupy specific regions of space and time, and that nothing other than a specific particular can be that particular. Examples are concrete persons such as George W. Bush Jr. and George W. Bush’s heart. Some particulars, such as each of four tanks in a specific squadron, may exactly look the same, but they are still distinct particulars. One can be destroyed, while the other three remain intact. For particulars of specific interest, such as persons, ships, and hurricanes, proper names are used to mark the importance of their individual identity. For other particulars, such as cars or pieces of complex equipment, serial numbers are used for unique identification purposes. Portion of Reality Entity Particular Universal Defined class Representation Non-referring particular Information bearer Denotator IUI RT-tuple corresponds-to Configuration represents CUI UUI denotes denotes is about Representational unit denotes contains Figure 1: Reality and representations W. Ceusters and S. Manzoor / How to Track Absolutely Everything 20 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 33. Universals, in contrast, are such that they are (1) generic and (2) expressed in language by means of general terms such as ‘person’, ‘ship’, and ‘car’, and (3) represent structures or characteristics in reality which are exemplified in an open-ended collection of particulars in arbitrarily disconnected regions of space and time. Through yet other data types, RT makes explicitly the distinction between two sorts of particulars: those that are ‘information bearers’, and those that are not; the latter called ‘non-referring particulars’. Whereas non-referring particulars belong exclusively to the first level of reality – they are pure first-order entities – information bearers play a role in both levels 1 and 3. Examples of information bearers are a piece of paper containing a text about a person’s educational background, and a digital object, such as an image of a person in an information system. Information bearers are about something else, while non- referring particulars are not about something else. Information bearers can be about not only non-referring particulars, an example being the driving license card of a person which is about its driving rights, but also about other information bearers, an example being a textual description of a specific person’s driving license, stating, for instance, that the name of the driver is almost not readable. A copy of such a driving license can be at the same time about both the card and the rights enjoyed by the license holder. 4.1. Relations between information bearers and portions of reality RT distinguishes explicitly and formally between various relations that obtain between information bearers and the various types of portions of reality it is capable of describing. These relations are: • is-about, which obtains between an information bearer and a portion of reality, such as, for example, a book about George W. Bush Sr. (the book being an information bearer) being about parts of the life of George W. Bush Sr. and his environment (a combination of several configurations in which figure, besides George W. Bush Sr., various other entities such as his advisors, friends, trips, speeches, and so forth). • corresponds-to, which obtains between an RT-tuple and a configuration; • represents, which obtains between a specific subtype of information bearer, namely what we call a ‘representation’, and some further entity (or collection of entities). A representation is thus such that (1) the information it contains is about an entity, and not a configuration, external to the representation and (2) it stands for or represents that entity. Examples are an image, record, description or map of the United States. Note that a representation (e.g. a description such as ‘the man over there on the corner’) represents a given entity even though it leaves out many aspects of its target. • denotes, which obtains between data-elements expressed by means of a data type that we call ‘denotator’ (see further) and an entity. • contains, which obtains between information bearers and can be used to express what pieces of information of a specific data type are parts of other pieces of information. An example is a digital message which contains RT- tuples describing configurations of entities in which a specific person figures. W. Ceusters and S. Manzoor / How to Track Absolutely Everything 21 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 34. 4.2. Denotators A denotator is a representational unit which denotes directly an entity in its entirety without providing a description. An example of a denotator is the string ‘Bush’ in the sentence ‘President Bush visited Europe several times’ when, whether or not known to the reader of the sentence in question, the writer had in mind a particular Bush, whether George Bush Jr. or George Bush Sr. The sentence itself is an information bearer according to our terminology. Because many representations are built out of constituent sub-representations as their parts, in the way in which paragraphs are built out of sentences and sentences out of words, RT uses the data type called ‘representational unit’ to represent such smallest part. Examples are: icons, names, simple word forms, or the sorts of alphanumeric identifiers found in digital records. Note that many images are not composite representations since they are not built out of smallest representational units in the way in which molecules are built out of atoms (Pixels are not representational units in the sense defined.) [18]. RT distinguishes explicitly and formally between three types of denotators, referred to respectively as ‘IUI’, ‘UUI’ and ‘CUI’. An IUI – abbreviation for ‘Instance Unique Identifier’ – is a denotator in the form of a persistent, globally unique and singular identifier which denotes (or is believed to denote) a particular and which is managed in a referent tracking system. A UUI – for ‘Universal Unique Identifier’ is a denotator which denotes a universal within the context of a realism-based ontology. A CUI – abbreviation for ‘Concept Unique Identifier’ – is a denotator for entities of a type that is commonly and ambiguously called a ‘concept’ [17], but which in BFO is called a ‘defined class’, and defined as a subset of the extension of a universal which is such that the members of this subset exhibit an additional property which is (a) not shared by all instances of the universal, and (b) also might be exhibited by particulars which are not instances of that universal. 5. Referent Tracking System A referent tracking system (RTS) is a special kind of digital information system which keeps track of (1) what is the case in reality and (2) what is expressed in other information systems about what is believed to be the case in reality. It does this unambiguously by means of the data types just sketched – in the first place resorting to IUIs – using principles and methods that assure – modulo the occurrence of errors, the resolution of which is also covered by the RT paradigm – that an IUI is (1) persistent because once created in a RTS it is never deleted, (2) globally unique because an IUI denotes only one entity within an RTS, and (3) singular because within an RTS, there is only one IUI for a specific entity. Figure 2 shows the various components of an RTS and how an RTS can be used in association with external information systems and terminology (or ontology) servers. The direction of the arrows depicted therein shows the processing of service requests, the communication, however, being bi-directional to accommodate responses to the requests. W. Ceusters and S. Manzoor / How to Track Absolutely Everything 22 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 35. 5.1. Components of a referent tracking system Figure 2: Components of a referent tracking system Referent Tracking Server (Peers) Referent Tracking System Referent Tracking Data Access Server External Information System Reasoning Server Referent Tracking System User Interface(s) User User Terminology Server Vocabulary Thesaurus Nomenclature Concept System Realism-based Ontology or or or or Referent Tracking Data Store RTS Proxy Peer RTS Server Proxy Peer Internal Ontology IUI Component An RTS includes at least four types of components: (1) one or more referent tracking servers, (2) one or more referent tracking system user interfaces, (3) an RTS Proxy Peer, and (4) an RTS Server Proxy Peer. The components execute on one or more processors, computers or computing devices. Further, all of the components of an RTS can run on one computing unit; one or more components can run on one computing unit, while others run on one or more other computing units; or the components may be distributed among various computing units. Each referent tracking server includes a data access server [24], which manages service requests coming from an RTS Proxy Peer or RTS Server Proxy Peer and which performs data manipulation on the server’s main component: a referent tracking data store thereby assisted by a reasoning server. The latter performs various sorts of reasoning functions by combining data from the data store with information coming from external terminology servers. The type of reasoning that can be performed depends on whether the terminology server contains nomenclatures, vocabularies, thesauri, and so forth. The referent tracking server comes also with an internal ontology which is a repository dedicated, for instance, to store information obtained during the initialization process, access control information about authorized users and usages, and so forth. The referent tracking system user interfaces allow direct users of the RTS to perform (1) a variety of management functions such as registering new external information systems, configuring a referent tracking server, adding additional referent W. Ceusters and S. Manzoor / How to Track Absolutely Everything 23 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 36. tracking servers, and so forth, and (2) content functions such as running pattern- matching algorithms on the data in the referent tracking data store to detect inconsistencies, invoke triggers and alerts, perform population-based studies, and so forth. 5.2. Layered architecture Figure 3 provides further details regarding the four-layered architecture of a RTS. The outer layer is a client side layer which connects to a RTS client which is typically a third party information system or a middleware component. The latter send a query to a Proxy Peer in the network layer that forwards the request to the appropriate RTS server in the network. During execution of the query, the RTS server calls the services of the RTS core API to retrieve the results from the Database Management System databases (DBMS) that constitute the data source layer. A referent tracking data store includes, for instance, two parts: an IUI-repository and a referent tracking database (RTDB). The IUI-repository includes, as explained in section 6, the A-tuples and D-tuples which provide meta-information about information about first-order entities. The IUI-repository thus manages the statements about the assignment of IUIs to particulars, and provides a central repository of IUIs to the RTS. The RTDB is a database of statements representing the detailed information about particulars, examples being ‘#IUI-1 instantiates the universal Person’ and ‘#IUI-1 has the name ‘John’’. The RTS Core layer implements the business logic of RT, namely, the insertion and retrieval of RT-tuples in any of its databases. The IUI-repository and RTDB components are implemented through a series of application programming interfaces (APIs). The IUI-repository includes services to search particular representations and to insert new ones in its corresponding DBMS. Similarly, the RTDB components provide API get methods to search and create methods to insert tuples in its database. Referent Tracking Data Access Server Referent Tracking System RTS Proxy Peer Information System RTS Services Factory RTS Services Factory Referent Tracking Data Store Referent Tracking Database Referent Tracking Database IUI Repository IUI Repository Database Managing System Database Managing System RTDB Tables IUI repository Tables IUI repository Tables RTS Services Server RTS Services Server Database Managing System Database Managing System Data source layer RTS core layer Network layer Client side layer Figure 3: Layered implementation of a referent tracking system W. Ceusters and S. Manzoor / How to Track Absolutely Everything 24 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 37. The IUI-repository and RTDB components are implemented independently of any specific DBMS (e.g. MYSQL, HSQL). DBMS support is controlled by DBMS specific driver components, such as for MYSQL and HSQL. Insertion services allow inserting a new RT tuple into the repository. The RT- tuples are inserted in a transaction, which is an information unit. As an example, entering a patient's blood pressure could involve a couple of RT statements which could include one or more RT-tuples. All tuples in a transaction are guaranteed to be committed in the data store. In case where either a system breaks down (by power failure or other means) or a user aborts the operation (e.g. a user closes/cancels the data entry screen while entering data), no partial information is stored in the data store. This service marks the start of a transaction for a specific session of a user. The RT paradigm does not allow any deletion operation in order to be able to always return to a state of the database as it was at a certain time in history. To prevent mistakes in creating new tuples in the IUI-repository, the tuples are cached right after the create operation. The client can remove or modify the tuples from the cache, as long as the commit service has not been called. 5.3. Networks of Referent Tracking Systems Since referent tracking is to make reference to entities in reality by means of singular and globally unique identifiers, an ideal setup is one in which only one RTS is used worldwide. More realistic, however, is the adoption of the RT paradigm in a step-wise fashion: each organization first installs its own RTS, and afterwards connects them in expanding networks. To support this evolution, as shown in , the RTS is built upon Peer to Peer (P2P) technology, enabling data sharing in such a way that a search query can be executed concurrently over distributed RTS servers (peers). In an RTS P2P network, a client thus sends a query to an RTS server which besides executing the query itself can forward it to other connected RTS servers for subsequent execution. Each peer then collects the results and sends them to the requesting peer. Finally, the RTS server who received the initial request returns the aggregated results to the client. Furthermore, an RTS P2P application is capable of database load sharing over multiple RTS server peers such that the network behaves as a singular database. This capability is useful in cases where a very large database cannot be hosted on a single machine, for instance because of computational limits. It includes also capabilities for discovering a new peer in a network, for authenticating users, and for ensuring secure communication. shows an example of an RTS network in which three organizations, A, B and C, are running their own RTS peers. The peers are installed so that they are not directly known outside their corresponding organization’s environment. In organization A, the Server Peers are alike in all respects and implement the objective of distributing a very large database load. When Information System A sends a search query to the RTS Proxy Peer within organization A, the latter forwards the query to all available Server Peers (A1, A2, …) in the organization which concurrently execute the query and return the results to the Proxy Peer that finally sends the results to the Information System. Each organization can form its own local group of servers whose membership is not known outside the organization. This protects against unauthorized access to the peers W. Ceusters and S. Manzoor / How to Track Absolutely Everything 25 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 38. in the group. Controlled public access to each organization’s data is offered through the Proxy Server peers. The separation of local peer advertisement within an organization from public (outside the host organization) contexts is the basis for the Referent Tracking Server C1 Referent Tracking Server C1 Referent Tracking System C RTS Proxy Peer RTS Server Proxy Peer Referent Tracking Server C2 Referent Tracking Server C2 Referent Tracking Server C3 Referent Tracking Server C3 … Referent Tracking Server B1 Referent Tracking Server B1 Referent Tracking System B RTS Proxy Peer RTS Server Proxy Peer Referent Tracking Server B2 Referent Tracking Server B2 Referent Tracking Server B3 Referent Tracking Server B3 … Referent Tracking Server A1 Referent Tracking Server A1 Referent Tracking System A RTS Proxy Peer RTS Server Proxy Peer Referent Tracking Server A2 Referent Tracking Server A2 Referent Tracking Server A3 Referent Tracking Server A3 … … Information System A Information System C Information System B implemented security layer. The peers which are known locally provide full access to the local database, and the peers which are known publicly provide very restricted access to the database (they might, for instance, allow only searches over certain sorts of RT-tuples as explained further). 5.4. Reasoning services Reasoning is a part of the RTS and its purpose is double. The first one is to prevent inconsistent data from being entered. By ‘inconsistent data’, we mean here data that cannot be true at the same time under the ontologies in whose terms the data are expressed. It is of course plausible that some analysts might be under the impression that, say ‘John is in Paris’ while others think that ‘John is in London’. That analysts think different things is not inconsistent, but clearly they cannot both be right. The second purpose for having reasoning services is to draw inferences during the execution of the search queries using the generic knowledge expressed in the terminology and ontology servers used to annotate the data and by exploiting the reasoners that operate on them. Various third party reasoners exist, some being specific to a particular knowledge source, some coming with a public DIG (Description Logic Implementation Group) interface for description logic representations, while others use directly OWL-DL (Web Ontology Language-Description Logics). Figure 4: Peer-to-Peer implementation of Referent Tracking Systems W. Ceusters and S. Manzoor / How to Track Absolutely Everything 26 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 39. In order to be able to deal with terminology servers and the various sorts of knowledge sources they offer (nomenclatures, thesauri, ontologies, ...), the RTS includes a Reasoning API which helps in sending reasoning queries uniformly to different terminology servers. The Reasoning API has an abstract class called OntologyConnector, which provides an interface to the external terminology systems by means of services. The interpretations of the OntologyConnector services are specific to a particular terminology server; therefore, a separate implementation of the OntologyConnector is required for each terminology server which is used to annotate the particulars in the RTS. Description logics are widely used for building ontologies. The reasoners for such ontologies may take from 1 second to a day to compute inferences over the ontology classes depending on their size and definitional complexity. Therefore, instead of always directly communicating with the reasoners for each ontology when a specific query is launched, the RTS is able to store these queries and the results that have been returned by these reasoners as an inference graph in a database [24]. Thus, because the execution time of the OntologyConnector services can range from milliseconds to minutes depending on the query execution time in the external terminology system, the OntologyConnector caches the results returned from these systems. The cache is stored, for instance, in a RDBMS. During the execution of any of the OntologyConnector services, it first searches in the cache. 6. Referent Tracking Data Elements: RT-tuples RT-tuples, although all corresponding to portions of reality, come in various flavors depending on the sort of information they contain. 6.1. A-tuples A-tuples correspond to the assignment by some agent of an IUI to a particular. For the typical case, that particular is a pure first-order entity such as a specific person or a specific building about which information is to be stored in the RT system. However, by storing tuples, the RT system itself acts as an agent that assigns IUIs to the tuples itself. Indeed, for each insertion of an A-tuple, there is a corresponding insertion of a D-tuple that contains information about the corresponding A-tuple. To prevent infinite regress, the assignment of these IUIs does not involve the generation of an additional A-tuple, but is implemented through the use of these tuple-IUIs as an internal annotation to the tuple itself. Three factors can be distinguished as structural elements involved in such an assignment act: (1) the generation of the relevant alphanumeric string, (2) its attachment to the relevant object, and (3) the publication of this attachment [20]. A-tuples are of the form < IUIp, IUIa, tap > where IUIp is the IUI of the particular in question, IUIa is the IUI of the author of the assignment act, and tap is a time-stamp indicating when the assignment was made. 6.2. D-tuples In light of the need or desire to resolve mistakes [25], RT includes the use of D-tuples, which are to be created whenever (1) a tuple other than a D-tuple is added to the RTS W. Ceusters and S. Manzoor / How to Track Absolutely Everything 27 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 40. Data Store, in which case it includes meta-data about by whom and at what time the corresponding tuple was deposited or (2) a tuple, including D-tuples, is declared invalid in the system, in which case it includes additional info concerning the type of mistake committed and the reason therefore. D-tuples are of the form < IUId, IUIT, td, E, C, S >, where: • IUIT is the IUI of the tuple about which the D-tuple contains information. • IUId: is the IUI of the entity annotating IUIT by means of this D-tuple, • E is either the symbol ‘I’ (for insertion) or any of the error type symbols as discussed further, • C is a symbol for the applicable reason for change as discussed further, • td is the time the tuple denoted by IUIT is inserted or ‘retired’, and • S is a list of IUIs denoting the tuples, if any, that replace the retired one. 6.3. PtoP-tuples Descriptions which express configurations amongst particulars have the form of PtoP – particular to particular – tuples. Here again a number of structural elements can be distinguished: (1) an authorized user observes one or more objects which have already been assigned IUIs in the referent tracking system (RTS) in hand, (2) the user recognizes or apprehends that these objects stand in a certain relation, which is represented in some realism-based ontology, (3) the user asserts that this relation holds and publishes this assertion by entering corresponding data which are then published in the referent tracking data store. This relationship data will then take the form of an ordered sextuple <IUIa, ta, r, IUIo, P, tr>, where • IUIa is the IUI of the author asserting that the relationship referred to by r holds between the particulars referred to by the IUIs listed in P; • ta is a time-stamp indicating when the assertion was made; • r is the denotator in IUIo of the relationship obtaining between the particulars referred to in P; • IUIo is the IUI of the ontology from which r is taken; • P is an ordered list of IUIs referring to the particulars between which r obtains; and • tr is a time-stamp representing the time at which the relationship was observed to obtain. P contains as many IUIs as are required by the arity of the relation r. In most cases, P will be an ordered pair which is such that r obtains between the particulars represented by its first and second IUIs when taken in this order. 6.4. PtoU-tuples Another type of information that can be provided about a particular concerns what universal within an ontology it instantiates. Here, too, time is relevant, since a particular, through development, growth or other changes, may cease to instantiate one universal and start to instantiate another: thus George W. Bush Sr. changed from foetus to newborn, and from child to adult. Descriptions of this type (which we will refer to as PtoU-tuples – for: particular to universal) are represented by ordered tuples of the form <IUIa, ta, inst, IUIo, IUIp, UUI, tr>, where W. Ceusters and S. Manzoor / How to Track Absolutely Everything 28 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 41. • IUIa is the IUI of the author asserting that IUIp is an instance (inst) of UUI; • ta is a time-stamp indicating when the assertion was made; • inst is the denotator in IUIo of the relationship of instantiation; • IUIo is the IUI of the realism-based ontology from which inst and UUI are taken; • IUIp is the IUI referring to the particular whose inst relationship with the universal denoted by UUI is asserted; • UUI is the denotator of the universal in IUIo with which IUIp enjoys the inst relationship; and • tr is a time-stamp representing the time at which the relationship was observed to obtain. Note that it is specified from which ontology inst and UUI are taken (and precisely which inst relationship in those cases where an ontology contains several variants). Such specifications not only ensure that the corresponding definitions can be accessed automatically, but also facilitate reasoning in the RTS Reasoning Server across ontologies that are interoperable with the ontology specified. 6.5. PtoC-tuples Whereas for PtoU-tuples their denotators of relationships and universals are taken from realism-based ontologies rather than from other knowledge repositories in terminology servers, PtoC-tuples do allow CUIs to be used instead of UUIs. Of course, the relationship to be used is not to be some variant of ‘inst’ since the standard definitions in use for ‘concept’ (such as ‘unit of knowledge’ or ‘unit of thought’) disallow most particulars from being declared as instances of concepts. PtoC-tuples (for particular to concept code) have the form <IUIa, ta, IUIc, IUIp, CUI, tr>, where: • IUIa is the IUI of the author asserting that terms associated to CUI may be used to describe IUIp; • ta is a time-stamp indicating when the assertion was made; • IUIc is the IUI of the concept-based system from which CUI is taken; • IUIp is the IUI referring to the particular which the author associates with CUI; • CUI is the CUI in the concept-system referred to by IUIc which the author associates with IUIp; and • tr is a time-stamp representing a time at which the author considers the association appropriate. Such tuples are to be interpreted as providing a facility equivalent to a simple index of terms in a work of scientific literature. 6.6. PtoU(-) – tuples Since the RT paradigm requires that only entities that exist or have existed are to be assigned an IUI, a capability is provided that deals with what is called ‘negative findings’ or ‘negative observations’ as captured in expressions such as: ‘no criminal history’, ‘membership of terrorist organization ruled out’, ‘absence of imminent danger’, and ‘attack prevented’. Such statements seem at first sight to present a problem for the referent tracking paradigm, since they imply that there are no entities in reality to which appropriate unique identifiers could be assigned. We therefore defined the relationship ‘p lacks u with respect to r at time t’ such that there obtains a relation W. Ceusters and S. Manzoor / How to Track Absolutely Everything 29 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 42. between the particular p and the universal u at time t, which is such that p stands to no instance of u in the relationship r at t [26, 27]. This ontological relation can be expressed by means of a ‘PtoU(-) tuple’ which is a lacks-counterpart of the PtoU-tuple and has the form <IUIa, ta, r, IUIo, IUIp, UUI, tr>, expressing that the particular referred to by IUIa asserts at time ta that the relation r of ontology IUIo does not obtain at time tr between the particular referred to by IUIp and any of the instances of the universal UUI at time tr. 6.7. PtoN-tuples Important particulars such as persons, ships, hurricanes, and so forth are often given proper names which function as denotators in reality outside the context of a referent tracking system. This sort of information is stored in an RTS by means of one or more ‘PtoN-tuples’ where ‘N’ stands for ‘name’. These tuples have the form < IUIa, ta, nt, n, IUIp, tr , IUIc >, where • IUIa is the IUI of the author asserting that n is a name of type nt used by IUIc to denote IUIp; • ta is a time-stamp indicating when the assertion was made; • IUIc is the IUI for the particular that uses the name n (this can be a person, a community of persons, an organization, an information system, ...); • IUIp is the IUI referring to the particular which the author associates with n; • n is the name which the author associates with IUIp; • nt is the nametype (examples being first name, last name, nick name, social security number, and so forth); and • tr is a time-stamp representing a time at which the author considers the association appropriate. 7. Discussion 7.1. Referent Tracking and action-oriented formalisms RT, at first sight, might look similar to other approaches. For instance, the need to track objects through time as they change, and to reason (and to have machines sometimes reason) over information that describes such changes, is what motivated calculi such as the situation calculus, the event calculus, and the fluent calculus, as well as some Knowledge Representation and Reasoning Systems. These approaches seek an efficient solution to the projection problem [28]: given an action theory that specifies the preconditions and effects of actions (including sensing), and a knowledge base about the initial state of the world, determine whether or not some condition holds after a given sequence of actions has been performed [29]. The situation calculus is a logic formalism that was first introduced by John McCarthy in 1963 [30] and since then underwent a few modifications [31]. The basic elements of situation calculus are: (1) actions that can be performed in the world, (2) fluents that describe the state of the world, each fluent thus being the representation of some property, and (3) situations. McCarthy and Hayes considered a situation to be ‘a complete state of the universe at an instant of time’ [32], a position which is also maintained in fluent calculus [33], whereas others redefined situations as finite W. Ceusters and S. Manzoor / How to Track Absolutely Everything 30 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 43. sequences of actions, thus a history of actions [31]. Event calculus does without situations, and uses only actions and fluents, whereby the latter are functions – rather than predicates as is the case in situation calculus – which can be used in predicates such as HoldsAt to state at what time which fluents hold [34]. RT differs in substantial ways from these logical formalisms. First of all, the goal of RT is not just to represent actions and changes, but all entities that exist in reality. Furthermore, these sorts of logics focus on computational aspects, but do not provide an integrated ontological characterization of entities such as actions, plans, and, because of their four-dimensionalist nature, for sure not of objects. It has been shown that it pays off to add more ontological rigor to formalisms such as situation calculus, for instance by using it only as one component for causal reasoning within a more elaborate, multi-component system [35]. RT, in contrast, is not in the first place a computational framework, but rather a representational one anchored in the realist view adhered to in Basic Formal Ontology (BFO) [23]. BFO distinguishes, for instance, continuants (such as George W. Bush) from occurrents (such as George W. Bush’s life or his last trip from Washington to New York). These distinctions, including BFO’s treatment of locations, positions and location schemes, was deemed essential in building a robot navigation model on top of situation calculus as embedded in Kuipers’ Spatial Semantic Hierarchy [36]. Relationships of the sort expressed by, for instance, RT’s PtoP- and PtoU-tuples hold only during certain time-periods [37, 38], and when they hold is expressed in the corresponding tuples themselves. In addition, PtoU-tuples express what universals a particular instantiates, thus also whether the entity described is an action or an object. Although no attempt has been made thus far, it seems plausible to assume that it is possible to express part of an RT database in terms of situation or event calculus. 7.2. Facts versus beliefs The requirements within RT that tuples must make direct and explicit reference to that what they are about, and that this can only be done for entities that exist or have existed, would seem to make it very difficult to represent uncertain, or possibly deceptive knowledge. One can wonder if, for example, an intercepted communication contains ‘Cain will strike down Abel’ and it is believed that ‘Cain’ and ‘Abel’ are code words for non-personal entities, whether this belief can be recorded in this system. Similar questions can be asked about things in the future: isn't it important for a representational framework to be able to state knowledge about future happenings and entities that might not exist until the future, such as tomorrow’s sunset or Al-Qaeda’s next attack? It is here that the distinction between three levels of reality as discussed in section 3.1 and the assignment of IUIs to RT-tuples themselves play a role. If a PtoP-tuple to which IUI-457 is assigned states that George W. Bush was president of the US in 2007, then the latter is taken to be a representation of reality – which of course may be a mistake – whereas IUI-457 is the proposition that the latter is the case. That this proposition is entertained (or not) by a specific person can be expressed by additional PtoP-tuples that relate the tuple in question to that person by referring also to adequate belief-related relations or processes depending on what sort of ontology is used. As in the case of action logics, RT itself does not come with a logic of beliefs, but from the representations, so we believe, secondary representations in terms of a belief logic can be generated. W. Ceusters and S. Manzoor / How to Track Absolutely Everything 31 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 44. For entities in the future, RT offers the possibility to reserve IUIs, rather than to assign IUIs [20]. Thus it is possible to assign an IUI to the plan to see and enjoy next Sunday’s sunset, whereas the detailed RT representation of that plan itself would contain a reserved IUI for that particular sunset. 7.3. Maintaining integrity There are several challenges in maintaining the representational integrity of an RT system, specifically with respect to the requirements that an IUI within an RTS should denote only one entity, and that there is only one IUI for a specific entity. If, for instance, one doesn’t know that ‘Usama bin Ladin’ and ‘Osama bin Laden’ denote the same individual, how could one possibly know to relate both names to the IUI denoting that individual? Here responsibility for faithful representation is shared between the user and the user interface. Whereas the former must devote enough effort to find out in each specific case what individual a name denotes, the latter, assisted by additional applications, must make it possible to reduce the effort required. Term comparison algorithms might be used to inform a user that a name similar to the one entered is already registered. Triggers and alerts can be implemented to warn a user that distinct individuals have the same name, and so forth. All this, however, does not guarantee that the right decision will be made in every case, and errors will very likely occur. So there have to be procedures to detect and correct mistakes. It is here that the D-tuples play an important role [25]. Easy to solve, once detected, are mistakes in which a particular has been assigned more than one IUI. In this case, only one of these IUIs would be used in future tuples, whereas all tuples in which the other IUIs are used will be replaced by tuples in which that one IUI will replace the redundant ones. This mechanism guarantees that it still remains known that during some period in the past, information concerning one particular was believed to be about two or more particulars. More work would be required in the opposite case, i.e. when the same IUI is used to denote distinct particulars. Here it might be necessary to perform a manual revision of the tuples in which that is used. To detect mistakes, the ontologies in whose terms RT-tuples are expressed can be used to guide integrity-checking routines that run over the RTDB. Because, for instance, persons (or any material continuant) cannot be at two distinct places in the same time, the presence of RT-tuples in the RTS that suggest this to be the case, indicates a mistake of the type ‘one IUI for distinct particulars’. Logically, because two distinct material continuants cannot occupy the same spatial region, any collection of RT-tuples representing that this would be the case must contain an error of the type ‘distinct IUIs for the same particular’. 7.4. RT and the Semantic Web The various types of tuples enumerated in section 6 are expressible using standard Semantic Web technologies, though with some additional formalisms implemented at the data-base storage level. This is indeed the approach that has been taken in implementing the system [24]. The Resource Description Framework (RDF) [39] was used as the basic representation language. Our RDF representations of the RT-tuples are treated as resources themselves: each resource is therefore prefixed with the RTS name space W. Ceusters and S. Manzoor / How to Track Absolutely Everything 32 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 45. URI and the prefix ‘rts:’ such that, for instance, the resource rts:IUI-1 is the same as http://org.buffalo.edu/RTS#IUI-1. To declare properties for resources, we used RDFS and mapped the RT-tuples to RDFS classes, thereby ensuring that the class names are identical to the template names, with the exception of PtoU-, which, because of restrictions in the RDFS naming conventions, has been mapped to PtoLackU. Our implementation of the RTS is accessible through Web services which are invoked through SOAP messages [40] containing the procedure information (procedure name, parameters and return type) and port type (location of the procedure). The RTS uses Axis for Java [41] to host the web services thereby taking advantage of the native support of the Web Services Definition Language (WSDL) [42] that Axis provides. The RTS has been build to be independent of any data source technology. To achieve this goal, we have defined the RTRepository class as an abstract Java class. This class provides all necessary services for managing the data based on the principles defined in the RT paradigm. To manage the RT data in a specific data source technology, an extension of the RTRepository for that specific technology is required. We have decided to develop the RTRepositorySesameImp class by extending the RTRepository such that it targets the SAIL Sesame API for manipulating RDF graphs as a data source [43]. Because the RT data are expressed in RDF, RDF query languages such as RQL [44], SPARQL [45] and SeRQL [43] can be used for retrieval. To this end, the RTRepository comes with the service ‘repository.query(querystring, language)’ which has an argument for the query string and a second one for the name of the query language in which the first argument is expressed. The SeRQL query language is implemented with the help of the Sesame SeRQL query language module, and the SPARQL query language is implemented with the help of the ARQ query module (a SPARQL processor for Jena) [46]. 8. Conclusion: meeting the new intelligence criteria When set up in appropriate ways, a network of referent tracking systems is able to meet all the requirements identified for the envisioned Globally Networked and Integrated Intelligence Enterprise (see section 1.3). The requirement to share intelligence data while still addressing the need to protect privacy, civil liberties, and sources and methods (C1), can be met by using the IUIs, typically the ones that stand proxy for persons, as pseudonyms. It would even be possible to go much further, for instance that all the information collected by credit card companies, banks, department stores, telecom providers and so forth would be pooled. Most citizens would find it unacceptable if that information were used for intelligence purposes without there being any reason to do so. But with the appropriate setup of IUIRepositories and RTDBs in such a way that, for instance, one specific agency has the means to link IUIs to persons, but otherwise no access to other RT-data, while other agencies would be able to do data-mining and pattern analysis on the pseudonymized data, no privacy or civil liberties would be violated. When analysts would detect suspicious patterns in the pseudonymized data pool, similar mechanisms as search warrants can be used to obtain re-identification of the data. The requirements to provide services across agencies, partners, and international borders for multiple mission use (C2) and to be able to adapt rapidly to changing needs W. Ceusters and S. Manzoor / How to Track Absolutely Everything 33 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 46. and new partners (C3) are supported by the possibility for referent tracking systems to cooperate in growing networks. The C4 requirement, i.e. to have security built into the data and environment using tags, together with the C5-requirement that access should be based on attributes that go beyond security classification, is met by the specific ways in which RT-tuples are set up: they contain in every case an indicator for the provenance of the data and all data are coded by means of ontologies or terminologies. Furthermore, each RT-tuple can be treated as a first-order entity, thereby receiving its own IUI, and that IUI can be used in other RT-tuples, for instance to describe to what type of entities or specific entities it may be disclosed. The same IUI can be used to track the flow of the data-element throughout the intelligence network. Data stewardship, finally, focusing on quality and reusability of data rather than, but not excluding, protection (C6) is a natural feature of the paradigm. One reason are the principles for IUI assignment which require that before an IUI is assigned to an entity, it should be checked whether that entity has already an IUI assigned to it. Mistakes will happen, of course, but they are traceable over time; if, for instance, when data accumulate, two IUIs start to appear repeatedly in the same configuration, then they may stand proxy for the same entity. Or, if the database at some stage contains a PtoP-tuple stating that the entity with IUIx was in some place at a given point in time, while in a completely different place a bit later, then it is likely, modulo other types of mistakes, that IUIx is denoting different things. A problem, at first sight, might be the amount of work required to represent information in this way. But here again, other types of software such as natural language processing applications, might assist. Furthermore, as shown in [47, 48], it is in many cases possible to translate structured information into a form that is RT- compatible automatically. We argue that the effort to make systems of this kind acceptable is not greater than the effort to bring about the change in mindset to realize Vision 2015. 9. References [1] Central Intelligence Agency. What is Intelligence? 2007 June 20, 2008 [cited 2008 August 12]; Available from: https://www.cia.gov/news-information/featured-story-archive/2007-featured-story- archive/what-is-intelligence.html [2] Reagan R. Executive Order 12333--United States intelligence activities. 1981. [3] United States Intelligence Community. The Intelligence Process. 2008 [cited 2008 August 12]; Available from: http://www.intelligence.gov/2-business.shtml [4] Travers R. A Blueprint for Survival; The Coming Intelligence Failure. Studies in Intelligence. 1997;Semiannual Edition, No. 1:35-43. [5] Chen H. Intelligence and Security Informatics for International Security. Information Sharing and Data Mining. New York: Springer-Verlag 2006. [6] Office of the Director of National Intelligence. United States Intelligence Community Information Sharing Strategy. 2008. [7] Hasselbring W. Information system integration. Communications of the ACM. 2000;43(6):33-8. [8] Office of Homeland Security. National Strategy for Homeland Security. Washington D.C.: Office of Homeland Security 2002. [9] H. Chen, R. Miranda, D. Zeng, T. Madhusudan, C. Demchak, Schroeder J. Intelligence and Security Informatics: Proceedings of the First Symposium on Intelligence and Security Informatics (ISI’03). New York: Springer-Verlag 2003. [10] Chen H, Wang F-Y, Zeng D. Intelligence and Security Informatics for Homeland Security: Information, Communication, and Transportation. IEEE Transactions on Intelligent Transportation Systems. 2004 December;5(4):329-41. W. Ceusters and S. Manzoor / How to Track Absolutely Everything 34 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 47. [11] Wang GA, Atabakhsh H, Petersen T, Chen H. Discovering Identity Problems: A Case Study. IEEE international conference on intelligence and security informatics. Atlanta, GA, USA Springer 2005. [12] Cybenko G, Giani A, Thompson: P. Cognitive Hacking: A Battle for the Mind. IEEE Computer. 2002;35(8):50-6. [13] Mercado S. A Venerable Source in a New Era: Sailing the Sea of OSINT in the Information Age. CIA Studies in Intelligence. 2004;48(3):45-55. [14] Office of the Director of National Intelligence. Office of the Director of National Intelligence. 2008 August 22 [cited 2008 August 28]; Available from: http://www.dni.gov/index.html [15] Office of the Director of National Intelligence. Vision 2015: A Globally Networked and Integrated Intelligence Enterprise. 2008. [16] Office of the Director of National Intelligence. The National Intelligence Strategy of The United States of America. Office of the Director of National Intelligence 2005. [17] Smith B. Beyond concepts: ontology as reality representation. Proceedings of the third international conference on formal ontology in information systems (FOIS 2004). Amsterdam: IOS Press 2004:73-84. [18] Smith B, Kusnierczyk W, Schober D, Ceusters W. Towards a Reference Terminology for Ontology Research and Development in the Biomedical Domain. KR-MED 2006, Biomedical Ontology in Action. Baltimore MD, USA 2006. [19] Bittner T, Smith B. A Theory of Granular Partitions. In: Duckham M, Goodchild MF, Worboy MF, eds. Foundations of Geographic Information Science. London: Taylor & Francis Books 2003:117-51. [20] Ceusters W, Smith B. Referent Tracking in Electronic Healthcare Records. In: Engelbrecht R, Geissbuhler A, Lovis C, Mihalas G, eds. Connecting Medical Informatics and Bio-Informatics Medical Informatics Europe 2005. Amsterdam: IOS Press 2005:71-6. [21] Ceusters W, Smith B. Referent Tracking for Digital Rights Management. International Journal of Metadata, Semantics and Ontologies. 2007;2(1):45-53. [22] Ceusters W, Smith B. Referent Tracking for Corporate Memories. In: Rittgen P, ed. Handbook of Ontologies for Business Interaction. New York and London: Idea Group Publishing 2007:34-46. [23] Grenon P, Smith B, Goldberg L. Biodynamic Ontology: Applying BFO in the Biomedical Domain. In: Pisanelli DM, ed. Ontologies in Medicine. Amsterdam: IOS Press 2004:20-38. [24] Manzoor S, Ceusters W, Rudnicki R. Implementation of a Referent Tracking System. International Journal of Healthcare Information Systems and Informatics. 2007;2(4):41-58. [25] Ceusters W. Dealing with Mistakes in a Referent Tracking System. In: KS H, ed. Proceedings of Ontology for the Intelligence Community 2007 (OIC-2007). Columbia MA 2007:5-8. [26] Ceusters W, Elkin P, Smith B. Referent Tracking: The Problem of Negative Findings. In: Hasman A, Haux R, Lei Jvd, Clercq ED, Roger-France F, eds. Studies in Health Technology and Informatics Ubiquity: Technologies for Better Health in Aging Societies - Proceedings of MIE2006. Amsterdam: IOS Press 2006:741-6. [27] Ceusters W, Elkin P, Smith B. Negative Findings in Electronic Health Records and Biomedical Ontologies: A Realist Approach. International Journal of Medical Informatics. 2007 March;76:326-33. [28] Reiter R. Knowledge in Action. Logical Foundations for Specifying and Implementing Dynamical Systems. Boston: MIT Press 2001. [29] Vassos S, Levesque H. Progression of Situation Calculus Action Theories with Incomplete Information. In: Veloso M, ed. Proceedings of IJCAI-07 2007. [30] McCarthy J. Situations, actions and causal laws. Stanford, CA: Stanford University Artificial Intelligence Laboratory; 1963. [31] Reiter R. The frame problem in the situation calculus: a simple solution (sometimes) and a completeness result for goal regression. In: Lifshitz V, ed. Artificial intelligence and mathematical theory of computation: papers in honour of John McCarthy. San Diego, CA, USA: Academic Press Professional, Inc 1991:359-80. [32] McCarthy J, Hayes PJ. Some philosophical problems from the standpoint of artificial intelligence. Machine Intelligence. 1969;4:463-502. [33] Thielscher M. Introduction to the Fluent Calculus. Electronic Transactions on Artificial Intelligence. 1998;2(3-4):179-92. [34] Kowalski R. Database updates in the event calculus. Journal of Logic Programming. 1992;12(1-2):121- 46. [35] Kuipers B. The spatial semantic hierarchy. Artificial Intelligence. 2000 May;119(1-2):191 - 233. [36] Bateman J, Farrar S. Modelling Models of Robot Navigation Using Formal Spatial Ontology. Spatial Cognition IV Reasoning, Action, and Interaction. Berlin / Heidelberg: Springer 2005:366-89. [37] Smith B, Ceusters W, Klagges B, Köhler J, Kumar A, Lomax J, et al. Relations in biomedical ontologies. Genome Biology. 2005;6(5):R46. [38] Smith B, Grenon P. The Cornucopia of Formal-Ontological Relations. Dialectica. 2004;58(3):279-96. [39] Manola F, Miller E. RDF Primer. 2004 [cited; Available from: http://www.w3.org/TR/rdf-primer/ W. Ceusters and S. Manzoor / How to Track Absolutely Everything 35 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 48. [40] Mitra N. SOAP Version 1.2 Part 0: Primer. W3C Recommendation 2003. [41] The Apache Software Foundation. Axis: A Webservices toolkit. 2005 [cited 25 January, 2007]; Available from: http://ws.apache.org/axis/ [42] Christensen E, Curbera F, Meredith G, Weerawarana S. Web Services Description Language (WSDL) 1.1. W3C Note 2001. [43] Broekstra J, Kampman A, Harmelen Fv. Sesame: A Generic Architecture for Storing and Querying RDF and RDF Schema. Lecture Notes in Computer Science - International Semantic Web Conference ISWC2002. Heidelberg: Springer 2002:54-68. [44] Foundation for Research and Technology – Hellas. The RDF Query Language (RQL). 2003 July 18 [cited 25 January 2007]; Available from: http://139.91.183.30:9090/RDF/RQL/ [45] Prud'hommeaux E, Seaborne A. SPARQL Query Language for RDF. W3C Working Draft 2006 October 4th [cited January 22, 2007]; Available from: http://www.w3.org/TR/rdf-sparql-query/ [46] RDF Data Access Working Group. ARQ - A SPARQL Processor for Jena. 2007 [cited 15th Febuary, 2007]; Available from: http://jena.sourceforge.net/ARQ/ [47] Rudnicki R, Ceusters W, Manzoor S, Smith B. What Particulars are Referred to in EHR Data? A Case Study in Integrating Referent Tracking into an Electronic Health Record Application. In: Teich JM, Suermondt J, C H, eds. American Medical Informatics Association 2007 Annual Symposium Proceedings, Biomedical and Health Informatics: From Foundations to Applications to Policy. Chicago, IL 2007:630-4. [48] Manzoor S, Ceusters W, Rudnicki R. A Middleware Approach to Integrate Referent Tracking in EHR Systems. In: Teich JM, Suermondt J, C H, eds. Proceedings of the American Medical Informatics Association 2007 Annual Symposium Biomedical and Health Informatics: From Foundations to Applications to Policy. Chicago IL 2007:503-7. W. Ceusters and S. Manzoor / How to Track Absolutely Everything 36 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 49. Chapter 3 Uses of Ontologies in Open Source Blog Mining Brian Ulicnya , Mieczyslaw M. Kokara,b , Christopher J. Matheusa a VIStology, Inc. b Northeastern University Abstract: The blogosphere provides a novel window into an important segment of public opinion, but its dynamic nature makes it an elusive medium to analyze and interpret in the aggregate, where it is most informative. We are developing a new open-source blog mining technology that employs ontologies to solve this problem by fusing the signals of the blogosphere and zeroing in on issues that are most likely to migrate offline. This technology is designed to enable analysts to anticipate the threats or opportunities these issues represent in a timely and efficient fashion. Keywords: Blog mining, Malaysia, human terrain, situation awareness Introduction Although much, perhaps even the majority, of what is discussed in the blogosphere is of little consequence and fleeting interest, blogs continue to emerge as powerful organizing mechanisms, giving momentum to ideas that shape public opinion and influence behavior. There are nearly 16 million active blogs [17] on the Internet with more launched every day, and bloggers have increasingly made an impact politically in a range of places and situations. For example, Malaysian bloggers have recently become quite assertive in confronting perceived corruption in their national government despite strict governmental control of the major media [1]. Although one must be careful not to extrapolate from the population of bloggers to the population as a whole, clearly blogs provide unparalleled access to an important segment of public opinion about events of the day. The perspective blog mining provides is much more complete than that provided by the letters to the editor section of a newspaper or magazine, if it has one. Even premier print newspapers such as the New York Times publish only 15 to 20 of the 1000 letters they receive daily in reaction to their reports [5]; by contrast, there are typically over 3,000 blog posts that cite New York Times stories for any particular day, including many posts not in English. By mining the unfiltered reactions of bloggers to the day’s events, we can clearly evaluate and quantify the reaction to the days’ news reports of a highly motivated group of users. This is particularly useful where the local press is tightly controlled. Ontologies and Semantic Technologies for Intelligence L. Obrst et al. (Eds.) IOS Press, 2010 © 2010 The authors and IOS Press. All rights reserved. doi:10.3233/978-1-60750-581-5-37 37 Ontologies and Semantic Technologies for Intelligence, edited by L. Obrst, et al., IOS Press, Incorporated, 2010. ProQuest Ebook Copyright © 2010. IOS Press, Incorporated. All rights reserved.
- 50. Other documents randomly have different content
- 54. The Project Gutenberg eBook of Manette Salomon
- 55. This ebook is for the use of anyone anywhere in the United States and most other parts of the world at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this ebook or online at www.gutenberg.org. If you are not located in the United States, you will have to check the laws of the country where you are located before using this eBook. Title: Manette Salomon Author: Edmond de Goncourt Jules de Goncourt Release date: September 11, 2020 [eBook #63179] Most recently updated: October 18, 2024 Language: French Credits: Produced by Carlo Traverso, Laurent Vogel and the Distributed Proofreading team at DP-test Italia. (This file was produced from images generously made available by The Internet Archive/Canadian Libraries.) *** START OF THE PROJECT GUTENBERG EBOOK MANETTE SALOMON ***
- 56. ROMANS DE EDMOND ET JULES DE GONCOURT MANETTE SALOMON NOUVELLE ÉDITION PARIS BIBLIOTHÈQUE-CHARPENTIER EUGÈNE FASQUELLE, ÉDITEUR 11, RUE DE GRENELLE, 11 1902 Tous droits réservés EUGÈNE FASQUELLE, ÉDITEUR, 11, RUE DE GRENELLE ŒUVRES DE EDMOND ET JULES DE GONCOURT
- 57. GONCOURT (Edmond de) La fille Élisa, 38e mille 1 vol. Les frères Zemganno, 8e mille 1 vol. La Faustin, 19e mille 1 vol. Chérie, 18e mille 1 vol. La Maison d'un artiste au XIXe siècle 2 vol. Les actrices du XVIIIe siècle: Mme Saint-Huberty 1 vol. — Mlle Clairon (3e mille) 1 vol. — La Guimard 1 vol. Les Peintres japonais: Outamaro.—Le Peintre des Maisons vertes, 4e mille 1 vol. —Hokousaï (peintre), (2e mille) 1 vol. GONCOURT (Jules de) Lettres, précédées d'une préface de H. Céard (3e mille) 1 vol. GONCOURT (Edmond et Jules de) En 18** 1 vol. Germinie Lacerteux 1 vol. Madame Gervaisais 1 vol. Renée Mauperin 1 vol. Manette Salomon 1 vol. Charles Demailly 1 vol. Sœur Philomène 1 vol. Quelques créatures de ce temps 1 vol. Pages retrouvées, avec une préface de G. Geffroy (3e mille) 1 vol. Idées et sensations 1 vol.
- 58. Préfaces et manifestes littéraires (3e mille) 1 vol. Théâtre (Henriette Maréchal.—La Patrie en danger) 1 vol. Portraits intimes du XVIIIe siècle. Études nouvelles d'après les lettres autographes et les documents inédits 1 vol. La Femme au XVIIIe siècle 1 vol. La duchesse de Châteauroux et ses sœurs 1 vol. Madame de Pompadour, nouvelle édition, revue et augmentée de lettres et documents inédits 1 vol. La Du Barry 1 vol. Histoire de Marie-Antoinette 1 vol. Sophie Arnould (Les actrices au XVIIIe siècle) 1 vol. Histoire de la Société française pendant la Révolution 1 vol. Histoire de la Société française pendant le Directoire 1 vol. L'Art du XVIIIe siècle. 1re série (Watteau.—Chardin.—Boucher.—Latour) 1 vol. 2e série (Greuze.—Les Saint-Aubin.—Gravelot.— Cochin) 1 vol. 3e série (Eisen.—Moreau-Debucourt.—Fragonard.— Prudhon) 1 vol. Gavarni. L'Homme et l'Œuvre 1 vol. Journal des Goncourt. Mémoires de la vie littéraire (9e mille). 9 vol. Paris.—L. Maretheux, imprimeur, 1, rue Cassette.—1215.
- 59. MANETTE SALOMON
- 60. I On était au commencement de novembre. La dernière sérénité de l'automne, le rayonnement blanc et diffus d'un soleil voilé de vapeurs de pluie et de neige, flottait, en pâle éclaircie, dans un jour d'hiver. Du monde allait dans le Jardin des Plantes, montait au labyrinthe, un monde particulier, mêlé, cosmopolite, composé de toutes les sortes de gens de Paris, de la province et de l'étranger, que rassemble ce rendez-vous populaire. C'était d'abord un groupe classique d'Anglais et d'Anglaises à voiles bruns, à lunettes bleues. Derrière les Anglais, marchait une famille en deuil. Puis suivait, en traînant la jambe, un malade, un voisin du jardin, de quelque rue d'à côté, les pieds dans des pantoufles. Venaient ensuite: un sapeur, avec, sur sa manche, ses deux haches en sautoir surmontées d'une grenade;—un prince jaune, tout frais habillé de Dusautoy, accompagné d'une espèce d'heiduque à figure de Turc, à dolman d'Albanais;—un apprenti maçon, un petit gâcheur débarqué du Limousin, portant le feutre mou et la chemise bise. Un peu plus loin, grimpait un interne de la Pitié, en casquette, avec un livre et un cahier de notes sous le bras. Et presque à côté de lui, sur la même ligne, un ouvrier en redingote, revenant
- 61. d'enterrer un camarade au Montparnasse, avait encore, de l'enterrement, trois fleurs d'immortelle à la boutonnière. Un père, à rudes moustaches grises, regardait courir devant lui un bel enfant, en robe russe de velours bleu, à boutons d'argent, à manches de toile blanche, au cou duquel battait un collier d'ambre. Au-dessous, un ménage de vieilles amours laissait voir sur sa figure la joie promise du dîner du soir en cabinet, sur le quai, à la Tour d'argent. Et, fermant la marche, une femme de chambre tirait et traînait par la main un petit négrillon, embarrassé dans sa culotte, et qui semblait tout triste d'avoir vu des singes en cage. Toute cette procession cheminait dans l'allée qui s'enfonce à travers la verdure des arbres verts, entre le bois froid d'ombre humide, aux troncs végétants de moisissure, à l'herbe couleur de mousse mouillée, au lierre foncé et presque noir. Arrivé au cèdre, l'Anglais le montrait, sans le regarder, aux miss, dans le Guide; et la colonne, un moment arrêtée, reprenait sa marche, gravissant le chemin ardu du labyrinthe d'où roulaient des cerceaux de gamins fabriqués de cercles de tonneaux, et des descentes folles de petites filles faisant sauter à leur dos des cornets à bouquin peints en bleu. Les gens avançaient lentement, s'arrêtant à la boutique d'ouvrages en perles sur le chemin, se frôlant et par moments s'appuyant à la rampe de fer contre la charmille d'ifs taillés, s'amusant, au dernier tournant, des micas qu'allume la lumière de trois heures sur les bois pétrifiés qui portent le belvédère, clignant des yeux pour lire le vers latin qui tourne autour de son bandeau de bronze: Horas non numero nisi serenas. Puis, tous entrèrent un à un sous la petite coupole à jour. Paris était sous eux, à droite, à gauche, partout. Entre les pointes des arbres verts, là où s'ouvrait un peu le rideau des pins, des morceaux de la grande ville s'étendaient à perte de
- 62. vue. Devant eux, c'étaient d'abord des toits pressés, aux tuiles brunes, faisant des masses d'un ton de tan et de marc de raisin, d'où se détachait le rose des poteries des cheminées. Ces larges teintes étalées, d'un ton brûlé, s'assombrissaient et s'enfonçaient dans du noir-roux en allant vers le quai. Sur le quai, les carrés de maisons blanches, avec les petites raies noires de leurs milliers de fenêtres, formaient et développaient comme un front de caserne d'une blancheur effacée et jaunâtre, sur laquelle reculait, de loin en loin, dans le rouillé de la pierre, une construction plus vieille. Au delà de cette ligne nette et claire, on ne voyait plus qu'une espèce de chaos perdu dans une nuit d'ardoise, un fouillis de toits, des milliers de toits d'où des tuyaux noirs se dressaient avec une finesse d'aiguille une mêlée de faîtes et de têtes de maisons enveloppées par l'obscurité grise de l'éloignement, brouillées dans le fond du jour baissant; un fourmillement de demeures, un gâchis de lignes et d'architectures, un amas de pierres pareil à l'ébauche et à l'encombrement d'une carrière, sur lequel dominaient et planaient le chevet et le dôme d'une église, dont la nuageuse solidité ressemblait à une vapeur condensée. Plus loin, à la dernière ligne de l'horizon, une colline, où l'œil devinait une sorte d'enfouissement de maisons, figurait vaguement les étages d'une falaise dans un brouillard de mer. Là-dessus pesait un grand nuage, amassé sur tout le bout de Paris qu'il couvrait, une nuée lourde, d'un violet sombre, une nuée de Septentrion, dans laquelle la respiration de fournaise de la grande ville et la vaste bataille de la vie de millions d'hommes semblaient mettre comme des poussières de combat et des fumées d'incendie. Ce nuage s'élevait et finissait en déchirures aiguës sur une clarté où s'éteignait, dans du rose, un peu de vert pâle. Puis revenait un ciel dépoli et couleur d'étain, balayé de lambeaux d'autres nuages gris. En regardant vers la droite, on voyait un Génie d'or sur une colonne, entre la tête d'un arbre vert se colorant dans ce ciel d'hiver d'une chaleur olive, et les plus hautes branches du cèdre, planes, étalées, gazonnées, sur lesquels les oiseaux marchaient en sautillant comme sur une pelouse. Au delà de la cime des sapins, un peu balancés, sous lesquels s'apercevait nue, dépouillée, rougie, presque
- 63. carminée, la grande allée du jardin, plus haut que les immenses toits de tuile verdâtres de la Pitié et que ses lucarnes à chaperon de crépi blanc, l'œil embrassait tout l'espace entre le dôme de la Salpêtrière et la masse de l'Observatoire: d'abord, un grand plan d'ombre ressemblant à un lavi, d'encre de Chine sur un dessous de sanguine, une zone de tons ardents et bitumineux, brûlés de ces roussissures de gelée et de ces chaleurs d'hiver qu'on retrouve sur la palette d'aquarelle des Anglais; puis, dans la finesse infinie d'une teinte dégradée, il se levait un rayon blanchâtre, une vapeur laiteuse et nacrée, trouée du clair des bâtisses neuves, et où s'effaçaient, se mêlaient, se fondaient, en s'opalisant, une fin de capitale, des extrémités de faubourgs, des bouts de rues perdues. L'ardoise des toits pâlissait sous cette lueur suspendue qui faisait devenir noires, en les touchant, les fumées blanches dans l'ombre. Tout au loin, l'Observatoire apparaissait, vaguement noyé dans un éblouissement, dans la splendeur féerique d'un coup de soleil d'argent. Et à l'extrémité de droite, se dressait la borne de l'horizon, le pâté du Panthéon, presque transparent dans le ciel, et comme lavé d'un bleu limpide. Anglais, étrangers, Parisiens, regardaient de là-haut de tous côtés; les enfants étaient montés, pour mieux voir, sur le banc de bronze, quand quatre jeunes gens entrèrent dans le belvédère. —Tiens! l'homme de la lorgnette n'y est pas,—fit l'un en s'approchant de la lunette d'approche fixée par une ficelle à la balustrade. Il chercha le point, braqua la lunette:—Ça y est! attention!—se retourna vers le groupe d'Anglais qu'il avait derrière lui, dit à une des Anglaises:—Milady, voilà! confiez-moi votre œil… Je n'en abuserai pas! Approchez, mesdames et messieurs! Je vais vous faire voir ce que vous allez voir! et un peu mieux que ce préposé aux horizons du Jardin des Plantes qui a deux colonnes torses en guise de jambes… Silence! et je commence!… L'Anglaise, dominée par l'assurance du démonstrateur, avait mis l'œil à la lorgnette. —Messieurs! c'est sans rien payer d'avance, et selon les moyens des personnes!… Spoken here! Time is money! Rule Britannia! All
- 64. right! Je vous dis ça, parce qu'il est toujours doux de retrouver sa langue dans la bouche d'un étranger… Paris! messieurs les Anglais, voilà Paris! C'est ça!… c'est tout ça… une crâne ville!… j'en suis, et je m'en flatte! Une ville qui fait du bruit, de la boue, du chiffon, de la fumée, de la gloire… et de tout! du marbre en carton-papier, des grains de café avec de la terre glaise, des couronnes de cimetière avec de vieilles affiches de spectacle, de l'immortalité en pain d'épice, des idées pour la province, et des femmes pour l'exportation! Une ville qui remplit le monde… et l'Odéon, quelquefois! Une ville où il y a des dieux au cinquième, des éleveurs d'asticots en chambre, et des professeurs de thibétain en liberté! La capitale du Chic, quoi! Saluez!… Et maintenant ne bougeons plus! Ça? milady, c'est le cèdre, le vrai du Liban, rapporté d'un chœur d'Athalie, par M. de Jussieu, dans son chapeau!… Le fort de Vincennes! On compte deux lieues, mes gentlemen! On a abattu le chêne sous lequel Saint Louis rendait la justice, pour en faire les bancs de la cour de Cassation… Le château a été démoli, mais on l'a reconstruit en liége sous Charles X: c'est parfaitement imité, comme vous voyez… On y voit les mânes de Mirabeau, tous les jours de midi à deux heures, avec des protections et un passe-port… Le Père- Lachaise! le faubourg Saint-Germain des morts: c'est plein d'hôtels… Regardez à droite, à gauche… Vous avez devant vous le monument à Casimir Périer, ancien ministre, le père de M. Guizot… La colonne de Juillet, suivez! bâtie par les prisonniers de la Bastille pour en faire une surprise à leur gouverneur… On avait d'abord mis dessus le portrait de Louis-Philippe, Henri IV avec un parapluie; on l'a remplacé par cette machine dorée: la Liberté qui s'envole; c'est d'après nature… On a dit qu'on la muselait dans les chaleurs, à l'anniversaire des Glorieuses: j'ai demandé au gardien, ce n'est pas vrai… Regardez bien, mylady, il y a un militaire auprès de la Liberté: c'est toujours comme ça en France… Ça? c'est rien, c'est une église… Les buttes Chaumont… Distinguez le monde… On reconnaîtrait ses enfants naturels!… Maintenant, mylady, je vais vous la placer à Montmartre… La tour du télégraphe… Montmartre, mons martyrum… d'où vient la rue des Martyrs, ainsi nommée parce qu'elle est remplie de peintres qui s'exposent volontairement aux
- 65. bêtes chaque année, à l'époque de l'Exposition… Là-dessous, les toits rouges? ce sont les Catacombes pour la soif, l'Entrepôt des vins, rien que cela, mademoiselle!… Ce que vous ne voyez pas après, c'est simplement la Seine, un fleuve connu et pas fier, qui lave l'Hôtel-Dieu, la Préfecture de Police, et l'Institut!… On dit que dans le temps il baignait la Tour de Nesle… Maintenant, demi-tour à droite, droite alignement! Voilà Sainte Geneviève… A côté, la tour Clovis… c'est fréquenté par des revenants qui y jouent du cor de chasse chaque fois qu'il meurt un professeur de Droit comparé… Ici, c'est le Panthéon… le Panthéon, milady, bâti par Soufflot, pâtissier… C'est, de l'aveu de tous ceux qui le voient, un des plus grands gâteaux de Savoie du monde… Il y avait autrefois dessus une rose: on l'a mise dans les cheveux de Marat quand on l'y a enterré… L'arbre des Sourds-et-Muets… un arbre qui a grandi dans le silence… le plus élevé de Paris… On dit que quand il fait beau, on voit de tout en haut la solution de la question d'Orient… Mais il n'y a que le ministre des affaires étrangères qui ait le droit d'y monter!… Ce monument égyptien? Sainte-Pélagie, milady… une maison de campagne, élevée par les créanciers en faveur de leurs débiteurs… Le bâtiment n'a rien de remarquable que le cachot où M. de Jouy, surnommé «l'Homme au masque de coton», apprivoisait des hexamètres avec un flageolet… Il y a encore un mur teint de sa prose!… La Pitié… un omnibus pour les pékins malades, avec correspondance pour le Montparnasse, sans augmentation de prix, les dimanches et fêtes… Le Val-de-Grâce, pour MM. les militaires… Examinez le dôme, c'est d'un nommé Mansard, qui prenait des casques dans les tableaux de Lebrun pour en coiffer ses monuments… Dans la cour, il y a une statue élevée par Louis XIV au baron Larrey… L'Observatoire… Vous voyez, c'est une lanterne magique… il y a des Savoyards attachés à l'établissement pour vous montrer le Soleil et la Lune… C'est là qu'est enterré Mathieu Laensberg, dans une lorgnette… en long… Et ça… la Salpêtrière, milady, où l'on enferme les femmes plus folles que les autres! Voilà!… Et maintenant, à la générosité de la société! —lança le démonstrateur de Paris.
- 66. Il ôta son chapeau, fit le tour de l'auditoire, dit merci à tout ce qui tombait au fond de sa vieille coiffe, aux gros sous comme aux pièces blanches, salua et se sauva à toutes jambes, suivi de ses trois compagnons qui étouffaient de rire en disant:—Cet animal d'Anatole! Au cèdre, devant un vieux curé qui lisait son bréviaire, assis sur le banc contre l'arbre, il s'arrêta, renversa ce qu'il y avait dans son chapeau sur les genoux du prêtre, lui jeta:—Monsieur le curé, pour vos pauvres! Et le curé, tout étonné de cet argent, le regardait encore dans le creux de sa pauvre soutane, que le donneur était déjà loin.
- 68. II A la porte du Jardin des Plantes, les quatre jeunes gens s'arrêtèrent. —Où dine-t-on?—dit Anatole. —Où tu voudras,—répondirent en chœur les trois voix. —Qu'est-ce qui en a?—reprit Anatole. —Moi, je n'ai pas grand'chose,—dit l'un. —Moi, rien,—dit l'autre. —Alors ce sera Coriolis…—fit Anatole en s'adressant au plus grand, dont la mise élégante contrastait avec le débraillé des autres. —Ah! mon cher, c'est bête… mais j'ai déjà mangé mon mois… je suis à sec… Il me reste à peine de quoi donner à la portière de Boissard pour la cotisation du punch… —Quelle diable d'idée tu as eue de donner tout cet argent à ce curé!—dit à Anatole un garçon aux longs cheveux. —Garnotelle, mon ami,—répondit Anatole,—vous avez de l'élévation dans le dessin… mais pas dans l'âme!… Messieurs, je vous offre à dîner chez Gourganson… J'ai l'œil… Par exemple, Coriolis, il ne faut pas t'attendre à y manger des pâtés de harengs de Calais truffés comme à ta société du vendredi… Et se tournant vers celui qui avait dit n'avoir rien: —Monsieur Chassagnol, j'espère que vous me ferez l'honneur…
- 69. On se mit en marche. Comme Garnotelle et Chassagnol étaient en avant, Coriolis dit à Anatole, en lui désignant le dos de Chassagnol: —Qu'est-ce que c'est, ce monsieur-là, hein? qui a l'air d'un vieux fœtus… —Connais pas… mais pas du tout… Je l'ai vu une fois avec des élèves de Gleyre, une autre fois avec des élèves de Rude… Il dit des choses sur l'art, au dessert, il m'a semblé… Très-collant… Il s'est accroché à nous depuis deux ou trois jours… Il va où nous mangeons… Très-fort pour reconduire, par exemple… Il vous lâche à votre porte à des heures indues… Peut-être qu'il demeure quelque part, je ne sais pas où… Voilà! Arrivés à la rue d'Enfer, les quatre jeunes gens entrèrent par une petite allée dans une arrière-salle de crêmerie. Dans un coin, un gros gaillard noir et barbu, coiffé d'un grand chapeau gris, mangeait sur une petite table. —Ah! l'homme aux bouillons…—fit Anatole en l'apercevant. —Ceci, monsieur,—dit-il à Chassagnol,—vous représente… le dernier des amoureux!… un homme dans la force de l'âge, qui a poussé la timidité, l'intelligence, le dévouement et le manque d'argent jusqu'à fractionner son dîner en un tas de cachets de consommé… ce qui lui permet de considérer une masse de fois dans la journée l'objet de son culte, mademoiselle ici présente… Et d'un geste, Anatole montra mademoiselle Gourganson qui entrait, apportant des serviettes. —Ah! tu étais né pour vivre au temps de la chevalerie, toi! Laisse donc, je connais les femmes… j'avance joliment tes affaires, va, farceur!—et il donna un amical renfoncement au jeune homme barbu qui voulut parler, bredouilla, devint pourpre, et sortit. Le crêmier apparut sur le seuil: —Monsieur Gourganson! monsieur Gourganson!—cria Anatole,— votre vin le plus extraordinaire… à 12 sous!… et des bifteacks… des
- 70. vrais!… pour monsieur…—il indiqua Coriolis—qui est le fils naturel de Chevet… Allez! —Dis donc, Coriolis,—fit Garnotelle,—ta dernière académie… j'ai trouvé ça bien… mais très-bien… —Vrai?… vois-tu, je cherche… mais la nature!… faire de la lumière avec des couleurs… —Qui ne la font jamais…—jeta Chassagnol.—C'est bien simple, faites l'expérience… Sur un miroir posé horizontalement, entre la lumière qui le frappe et l'œil qui le regarde, posez un pain de blanc d'argent: le pain de blanc, savez-vous de quelle couleur vous le verrez? D'un gris intense, presque noir, au milieu de la clarté lumineuse… Coriolis et Garnotelle regardèrent après cette phrase, l'homme qui l'avait dite. —Qu'est-ce que c'est que ça?—Anatole, en cherchant dans sa poche du papier à cigarette, venait de retrouver une lettre.—Ah! l'invitation des élèves de Chose… une soirée où l'on doit brûler toutes les critiques du Salon dans la chaudière des sorcières de Macbeth… Il est bon, le post-scriptum: «Chaque invité est tenu d'apporter une bougie…» Et coupant une conversation sur l'École allemande qui s'engageait entre Chassagnol et Garnotelle:—Est-ce que vous allez nous embêter avec Cornélius?… Les Allemands! la peinture allemande!… Mais on sait comment ils peignent les Allemands… Quand ils ont fini leur tableau, ils réunissent toute leur famille, leurs enfants, leurs petits enfants… ils lèvent religieusement la serge verte qui recouvre toujours leur toile… Tout le monde s'agenouille… Prière sur toute la ligne… et alors ils posent le point visuel… C'est comme ça! C'est vrai comme… l'histoire! —Es-tu bête!—dit Coriolis à Anatole.—Ah ça! dis donc, tes bifteacks, pour des bifteacks soignés…
- 71. Welcome to Our Bookstore - The Ultimate Destination for Book Lovers Are you passionate about books and eager to explore new worlds of knowledge? At our website, we offer a vast collection of books that cater to every interest and age group. From classic literature to specialized publications, self-help books, and children’s stories, we have it all! Each book is a gateway to new adventures, helping you expand your knowledge and nourish your soul Experience Convenient and Enjoyable Book Shopping Our website is more than just an online bookstore—it’s a bridge connecting readers to the timeless values of culture and wisdom. With a sleek and user-friendly interface and a smart search system, you can find your favorite books quickly and easily. Enjoy special promotions, fast home delivery, and a seamless shopping experience that saves you time and enhances your love for reading. Let us accompany you on the journey of exploring knowledge and personal growth! ebookgate.com