Materializing the Web of Linked Data

Materializing the Web
of Linked Data
NIKOLAOS KONSTANTINOU
DIMITRIOS-EMMANUEL SPANOS

Book
“Materializing the Web of Linked Data”
Nikolaos Konstantinou and Dimitrios-Emmanuel Spanos
© Springer, 2015
ISBN
◦978-3-319-16073-3 (Print)
◦978-3-319-16074-0 (Online)
DOI
◦10.1007/978-3-319-16074-0
Materializing the Web of Linked Data

Contents
1. Introduction: Linked Data and the Semantic Web
2. Technical Background
3. Deploying Linked Open Data: Methodologies and
Software Tools
4. Creating Linked Data from Relational Databases
5. Generating Linked Data in Real-time from Sensor Data
Streams
6. Conclusions: Summary and Outlook

Chapter 1
Introduction
Linked Data and the Semantic Web

Outline
Introduction
Preliminaries
The Linked Open Data Cloud
Chapter 1 Materializing the Web of Linked Data 5

The Origin of the Semantic Web
Semantic Web
◦ Term primarily coined by Tim Berners-Lee
◦ For the Web to be understandable both by humans and software,
it should incorporate its meaning
◦ Its semantics
Linked Data implements the Semantic Web vision

Why a Semantic Web? (1)
Information management
◦ Sharing, accessing, retrieving, consuming information
◦ Increasingly tedious task
Search engines
◦ Rely on Information Retrieval techniques
◦ Keyword-based searches
◦ Do not release the information potential
◦ Large quantity of existing data on the web is not stored in HTML
◦ The Deep Web

Content meaning should be taken into account
◦ Structure to the meaningful Web page content
◦ Enable software agents to
◦ Roam from page to page
◦ Carry out sophisticated tasks for users
Semantic Web
◦ Complement, not replace current Web
◦ Transform content into an exploitable source of knowledge

Web of Data
◦ An emerging web of interconnected published datasets in the form
of Linked Data
◦ Implements the Semantic Web vision

The Need for Adding Semantics (1)
Semantics
◦ From the Greek word σημαντικός
◦ Pronounced si̱mantikós, means significant
◦ Term typically used to denote the study of meaning
◦ Using semantics
◦ Capture the interpretation of a formal or natural language
◦ Enables entailment, application of logical consequence
◦ About relationships between the statements that are expressed in this language

Syntax
◦ The study of the principles and processes by which sentences can
be formed
◦ Also of Greek origin
◦ συν and τάξις
◦ Pronounced sin and taxis
◦ Mean together and ordering, respectively

Two statements can be syntactically different but
semantically equivalent
◦ Their meaning (semantics) is the same, the syntax is different
◦ More than one ways to state an assumption

Search engines rely on keyword matches
◦ Keyword-based technologies
◦ Have come a long way since the dawn of the internet
◦ Will return accurate results
◦ Based on keyword matches
◦ Extraction and indexing of keywords contained in web pages
◦ But
◦ More complex queries with semantics still partially covered
◦ Most of the information will not be queried

Traditional search engines
Rely on keyword matches
Low precision
◦ Important results may not be fetched, or
◦ Ranked low
◦ No exact keyword matches

Keyword-based searches
Do not return usually the desired results
Despite the amounts of information
Typical user behavior
◦ Change query rather than navigate beyond the first page of the
search results

The Semantic Web (1)
Tackle these issues
◦ By offering ways to describe of Web resources
Enrich existing information
◦ Add semantics that specify the resources
◦ Understandable both by humans and computers
Information easier to discover, classify and sort
Enable semantic information integration
Increase serendipitous discovery of information
Allow inference

The Semantic Web (2)
Example
◦ A researcher would not have to visit every result page and
distinguish the relevant from the irrelevant results
◦ Instead, retrieve a list of more related, machine-processable
information
◦ Ideally containing links to more relevant information

Outline
Introduction
Preliminaries

Data-Information-Knowledge (1)
Data
Information
◦ “facts about a situation, person, event, etc.”
Smallest information particle
◦ The bit
◦ Replies with a yes or no (1 or 0)
◦ Can carry data, but in the same time, it can carry the result of a process
◦ For instance whether an experiment had a successful conclusion or not
Information, especially facts or numbers, collected to be examined
and considered and used to help decision-making, or information in
an electronic form that can be stored and used by a computer

E.g. “temperature” in a relational database could be
◦ The information produced after statistical analysis over numerous
measurements
◦ By a researcher wishing to extract the average value in a region over the
years
◦ A sensor measurement
◦ Part of the data that will contribute in drawing conclusions regarding climate
change

Knowledge
◦ Key term: “by experience or study”
◦ Implies processing of the underlying information
Understanding of or information about a subject that
you get by experience or study, either known by one
person or by people generally

No clear lines between them
◦ Term that is used depends on the respective point of view
When we process collected data in order to extract meaning, we
create new information
Addition of semantics
◦ Indispensable in generating new knowledge
◦ Semantic enrichment of the information
◦ Makes it unambiguously understood by any interested party (by people generally)
Data → Information → Knowledge

Heterogeneity
Distributed data sources
◦ Interconnected, or not
◦ Different models and schemas
◦ Differences in the vocabulary
◦ Syntactic mismatch
◦ Semantic mismatch

Interoperability (1)
Interoperable
◦ Two systems in position to successfully exchange information
3 non-mutually exclusive approaches
◦ Mapping among the concepts of each source
◦ Intermediation in order to translate queries
◦ Query-based
Protocols and standards (recommendations) are crucial
◦ E.g. SOAP, WSDL, microformats

Interoperability (2)
Key concept: Schema
◦ Word comes from the Greek word σχήμα
◦ Pronounced schíma
◦ Means the shape, the outline
◦ Can be regarded as a common agreement regarding the interchanged data
◦ Data schema
◦ Defines how the data is to be structured

Information Integration (1)
Combine information from heterogeneous systems, sources of
storage and processing
◦ Ability to process and handle it as a whole
Global-As-View
◦ Every element of the global schema is expressed as a query/view over
the schemas of the sources
◦ Preferable when the source schemas are not subject to frequent
changes

Information Integration (2)
Local-As-View
◦ Every element of the local schemas is expressed as a query/view over
the global schema
P2P
◦ Mappings among the sources exist but no common schema

Information Integration Architecture (1)
A source Π1 with schema S1
A source Π2 with a schema S2
…
A source Πν with source Sν
A global schema S in which the higher level queries are posed

Information Integration Architecture (2)
The goal in the information integration problem
◦ Submit queries to the global schema S
◦ Receive answers from S1, S2, …, Sν
◦ Without having to deal with or even being aware of the heterogeneity
in the information sources

Data Integration (1)
A data integration system
◦ A triple 𝐼 = 𝐺; 𝑆; 𝑀
◦ G is the global schema,
◦ S is the source schema and
◦ M is a set of mappings between G and S

Local-As-View
◦ Each declaration in M maps an element from the source schema S to
a query (a view) over the global schema G
Global-As-View
◦ Each declaration in M maps an element of the global schema G to a
query (a view) over the source schema S
The global schema G is a unified view over the heterogeneous
set of data sources

Same goal
◦ Unifying the data sources under the same common schema
Semantic information integration
◦ Addition of its semantics in the resulting integration scheme

Mapping
Using the definition of data integration systems, we can
define the concept of mapping
◦ A mapping m (member of M) from a schema S to a schema T
◦ A declaration of the form QS ⇝ QT
◦ QS is a query over S
◦ QT a query over T

Mapping vs. Merging
(Data) mapping
◦ A mapping is the specification of a mechanism
◦ The members of a model are transformed to members of another model
◦ The meta-model that can be the same, or different
◦ A mapping can be declared as a set of relationships, constraints, rules,
templates or parameters
◦ Defined during the mapping process, or through other forms that have not yet been defined
Merging
◦ Implies unifying the information at the implementation/storage layer

Annotation (1)
Addition of metadata to the data
Data can be encoded in any standard
Especially important in cases when data is not human-
understandable in its primary form
◦ E.g. multimedia

Annotation (2)
(Simple) annotation
◦ Uses keywords or other ad hoc serialization
◦ Impedes further processing of the metadata
Semantic annotation
◦ Describe the data using a common, established way
◦ Addition of the semantics in the annotation (i.e. the metadata)
◦ In a way that the annotation will be machine-processable
◦ In order to be able to infer additional knowledge

Problems with (Semantic) Annotation
Time-consuming
◦ Users simply do not have enough time or do not consider it important enough
in order to invest time to annotate their content
Familiarity required with both the conceptual and technical parts of
the annotation
◦ User performing the annotation must be familiar with both the technical as
well as the conceptual part
Outdated annotations
◦ A risk, especially in rapidly changing environments with lack of automation in
the annotation process

Automated Annotation
Incomplete annotation is preferable to absent
◦ E.g. in video streams
Limitations of systems performing automated annotation
◦ Limited recall and precision (lost or inaccurate annotations)

Metadata (1)
Data about data
Efficient materialization with the use of ontologies
Ontologies can be used to describe Web resources
◦ Adding descriptions about their semantics and their relations
◦ Aims to make resources machine-understandable
◦ Each (semantic) annotation can correspond to a piece of
information
◦ It is important to follow a common standard

Metadata (2)
Annotation is commonly referred to as “metadata”
Usually in a semi-structured form
◦ Semi-structured data sources
◦ Structure accompanies the metadata
◦ E.g. XML, JSON
◦ Structured data sources
◦ Structure is stored separately
◦ E.g. relational databases

Metadata (3)
Powerful languages
◦ Practically unlimited hierarchical structure
◦ Covers most of the description requirements that may occur
◦ Storage in separate files allows collaboration with communication protocols
◦ Files are independent from the environment in which they reside
◦ Resilience to technological evolutions
Negatives
◦ Expressivity of the description model
◦ Limited way of structuring the information

Metadata (4)
Inclusion of semantics
◦ Need for more expressive capabilities and terminology
Ontologies
◦ Offer a richer way of describing information
◦ E.g. defining relationships among concepts such as subclass/superclass,
mutually disjoint concepts, inverse concepts, etc.
◦ RDF can be regarded as the evolution of XML
◦ OWL enables more comprehensive, precise and consistent
description of Web Resources

Ontologies (1)
Additional description to an unclear model that aims to
further clarify it
Conceptual description model of a domain
◦ Describe its related concepts and their relationships
Can be understood both by human and computer
◦ A shared conceptualization of a given specific domain
Aim at bridging and integrating multiple and heterogeneous
digital content on a semantic level

Ontologies (2)
In Philosophy, a systematic recording of “Existence”
For a software system, something that “exists” is something
that can be represented
◦ A set of definitions that associate the names of entities in the
universe of discourse with human-readable text describing the
meaning of the names
◦ A set of formal axioms that constrain the interpretation and well-
formed use of these terms
The specification of a conceptualization (Gruber 1995)

Ontologies (3)
Can be used to model concepts regardless to how general or specific
these concepts are
According to the degree of generalization
◦ Top-level ontology
◦ Domain ontology
◦ Task ontology
◦ Application ontology
Content is made suitable for machine consumption
◦ Automated increase of the system knowledge
◦ Logical rules to infer implicitly declared facts

Reasoners (1)
Software components
Validate consistency of an ontology
◦ Perform consistency checks
◦ Concept satisfiability and classification
Infer implicitly declared knowledge
◦ Apply simple rules of deductive reasoning

Reasoners (2)
Basic reasoning procedures
◦ Consistency checking
◦ Concept satisfiablility
◦ Concept subsumption
◦ Instance checking (Realization)
Properties of special interest
◦ Termination
◦ Soundness
◦ Completeness

Reasoners (3)
Research in Reasoning
◦ Tradeoff between
◦ Expressiveness of ontology definition languages
◦ Computational complexity of the reasoning procedure
◦ Discovery of efficient reasoning algorithms
Commercial
◦ E.g. RacerPro
Free of charge
◦ E.g. Pellet, FaCT++

Knowledge Bases
Terminological Box (TBox)
◦ Concept descriptions
◦ Intensional knowledge
◦ Terminology and vocabulary
◦Assertional Box (ABox)
◦ The real data
◦ Extensional knowledge
◦ Assertions about named individuals in
terms of the TBox vocabulary
TBox
ABox
Description
Language
Reasoning
Knowledge Base
Application Programs Rules

Knowledge Bases vs. Databases (1)
A naïve approach:
◦ TBox ≡ Schema of the relational database
◦ ABox ≡ Schema instance
However, things are more complex than that

Knowledge Bases vs. Databases (2)
Relational model
◦ Supports only untyped relationships among relations
◦ Does not provide enough features to assert complex relationships among data
◦ Used in order to manipulate large and persistent models of relatively simple
data
Ontological scheme
◦ Allows more complex relationships
◦ Can provide answers about the model that have not been explicitly stated to
it, with the use of a reasoner
◦ Contains fewer but more complex data

Closed vs. Open World Assumption (1)
Closed world assumption
◦Relational databases
◦ Everything that has not been stated as true is false
◦ What is not currently known to be true is false
◦ A null value about a subject’s property denotes the non-existence
◦ A NULL value in the isCapital field of a table Cities claims that the city is not a capital
◦ The database answers with certainty
◦ A query “select cities that are capitals” will not return a city with a null value at a supposed
boolean isCapital field

Closed vs. Open World Assumption (2)
Open world assumption
◦ Knowledge Bases
◦ A query can return three types of answers
◦ True, false, cannot tell
◦ Information that is not explicitly declared as true is not necessarily false
◦ It can also be unknown
◦ Lack of knowledge does not imply falsity
◦ A question “Is Athens a capital city?” in an appropriate schema will
return “cannot tell” if the schema is not informed
◦ A database schema would return false, in the case of a null value

Monotonicity
A feature present in Knowledge Bases
A system is considered monotonic when new facts do not
discard existing ones
First version of SPARQL did not include update/delete
functionality
◦ Included in SPARQL 1.1
◦ Recommendation describes that these functions should be supported
◦ Does not describe the exact behavior

Outline
Introduction
Preliminaries

The LOD Cloud (1)
Structured data
Open format
Available for everyone to use it
Published on the Web and connected using Web
technologies
Related data that was not previously linked
◦ Or was linked using other methods

The LOD Cloud (2)
Using URIs and RDF for this is goal is very convenient
◦ Data can be interlinked
◦ Create a large pool of data
◦ Ability to search, combine and exploit
◦ Navigate between different data sources, following RDF links
◦ Browse a potentially endless Web of connected data sources

The LOD Cloud (3)
Applications in many cases out of the academia
Technology maturity
Open state/government data
◦ data.gov (US)
◦ data.gov.uk (UK)
◦ data.gov.au (Australia)
◦ opengov.se (Sweden)
Open does not necessarily mean Linked

The LOD Cloud (4)
Published datasets span several domains of human
activities
◦ Much more beyond government data
◦ Form the LOD cloud
◦ Constantly increasing in terms of volume

The LOD Cloud (5)
Evolution
Geographic
Media
Publications
User-Generated
Content
Government
Cross-Domain
Life Sciences
Social Networking
2007 2009 2014

Chapter 2
Technical Background

Outline
Introduction
RDF and RDF Schema
Description Logics
Querying RDF data with SPARQL
Mapping relational data with R2RML
Other technologies
Ontologies
Datasets

Introduction
Linked Data
◦ A set of technologies
◦ Focused on the web
◦ Use the Web as a storage and communication layer
◦ Provide meaning to web content
Semantics
◦ Added value when part of a larger context
◦ Data modeled as a graph
Technologies fundamental to the web

HTTP – HyperText Transfer Protocol
An application protocol for the management and transfer of
hypermedia documents in decentralized information systems
Defines a number of request types and the expected actions
that a server should carry out when receiving such requests
Serves as a mechanism to
◦ Serialize resources as a stream of bytes
◦ E.g. a photo of a person
◦ Retrieve descriptions about resources that cannot be sent over network
◦ E.g. the person itself

URI – Uniform Resource Identifier
URL
◦ Identifies a document location
◦ Address of a document or other entity that can be found online
URI
◦ Provides a more generic means to identify anything that exists in the
world
IRI
◦ Internationalized URI
IRI ⊇ URI ⊇ URL

HTML – HyperText Markup Language
A markup language for the composition and presentation of
various types of content into web pages
◦ E.g. text, images, multimedia
Documents delivered through HTTP are usually expressed in
HTML
◦ HTML5
◦ Current recommendation
◦ Additional markup tags
◦ Extends support for multimedia and mathematical content

XML – eXtensible Markup Language
Allows for strict definition of the structure of information
◦ Markup tags
The RDF model also follows an XML syntax

Outline
Introduction
RDF and RDF Schema
Description Logics
Other technologies
Ontologies
Datasets

Modeling Data Using RDF Graphs
Ontologies in the Semantic Web
◦ Model a system’s knowledge
◦ Based on RDF
◦ Model the perception of the world as a graph
◦ OWL builds on top of RDF
RDF
◦ A data model for the Web
◦ A framework that allows representing knowledge
◦ Model knowledge as a directed and labeled graph

RDF (1)
The cornerstone of the Semantic Web
A common representation of web resources
First publication in 1999
Can represent data from other data models
◦ Makes it easy to integrate data from multiple heterogeneous sources
Main idea:
◦ Model every resource with respect to its relations (properties) to
other web resources

RDF (2)
Relations form triples
◦ First term: subject
◦ Second term: property
◦ Third term: object
RDF statements contain triples, in the form:
(resource, property, resource)
or (subject, property, object)
ex:staffMember
ex:hasWorkAddress
ex:hasCity ex:hasAddress ex:hasPostalCode
“London” “23, Houghton str.” “WC2A 2AE”

Namespaces (1)
Initially designed to prevent confusion in XML names
Allow document elements to be uniquely identified
Declared in the beginning of XML documents
xmlns:prefix="location".

Namespaces (2)
Prefix Namespace URI Description
rdf http://www.w3.org/1999/02/22-rdf-syntax-ns# The built-in RDF vocabulary
rdfs http://www.w3.org/2000/01/rdf-schema RDFS puts an order to RDF
owl http://www.w3.org/2002/07/owl OWL terms
xsd http://www.w3.org/2001/XMLSchema# The RDF-compatible XML Schema datatypes
dc http://purl.org/dc/elements/1.1/ The Dublin Core standard for digital object description
foaf http://xmlns.com/foaf/0.1 The FOAF network
skos http://www.w3.org/2004/02/skos/core# Simple Knowledge Organization System
void http://rdfs.org/ns/void# Vocabulary of Interlinked Datasets
sioc http://rdfs.org/sioc/ns# Semantically-Interlinked Online Communities
cc http://creativecommons.org/ns Creative commons helps expressing licensing information
rdfa http://www.w3.org/ns/rdfa RDFa

RDF Serialization
RDF is mainly destined for machine consumption
Several ways to express RDF graphs in machine-readable
(serialization) formats
◦ N-Triples
◦ Turtle
◦ N-Quads
◦ TriG
◦ RDF/XML
◦ JSON-LD
◦ RDFa
Turtle family
RDFa
JSON-LD
Supports Multiple Graphs
RDF/XML
Turtle
N-Triples
TriG
N-Quads

N-Triples Serialization
<http://www.example.org/bradPitt> <http://www.example.org/isFatherOf>
<http://www.example.org/maddoxJoliePitt>.
<http://www.example.org/bradPitt> <http://xmlns.com/foaf/0.1/name> "Brad Pitt".
<http://www.example.org/bradPitt> <http://xmlns.com/foaf/0.1/based_near> :_x.
:_x <http://www.w3.org/2003/01/geo/wgs84_pos#lat> "34.1000".
:_x <http://www.w3.org/2003/01/geo/wgs84_pos#long> "118.3333".

Turtle Serialization
PREFIX ex: <http://www.example.org/>.
PREFIX foaf: <http://xmlns.com/foaf/0.1/>.
PREFIX geo: <http://www.w3.org/2003/01/geo/wgs84_pos#>.
ex:bradPitt ex:isFatherOf ex:maddoxJoliePitt;
foaf:name "Brad Pitt";
foaf:based_near :_x.
:_x geo:lat "34.1000";
geo:long "118.3333".

TriG Serialization
PREFIX ex: <http://www.example.org/>.
PREFIX foaf: <http://xmlns.com/foaf/0.1/>.
PREFIX geo: <http://www.w3.org/2003/01/geo/wgs84_pos#>.
GRAPH <http://www.example.org/graphs/brad> {
ex:bradPitt ex:isFatherOf ex:maddoxJoliePitt;
foaf:name "Brad Pitt";
foaf:based_near :_x.
:_x geo:lat "34.1000";
geo:long "118.3333".
}

XML/RDF Serialization
<?xml version="1.0" encoding="utf-8"?>
<rdf:RDF xmlns:ex="http://www.example.org/"
xmlns:foaf="http://xmlns.com/foaf/0.1/"
xmlns:geo="http://www.w3.org/2003/01/geo/wgs84_pos#">
<rdf:Description rdf:about=" http://www.example.org/bradPitt">
<ex:isFatherOf rdf:resource=" http://www.example.org/maddoxJoliePitt"/>
<foaf:name>Brad Pitt</foaf:name>
<foaf:based_near rdf:nodeID="A0"/>
</rdf:Description>
<rdf:Description rdf:nodeID="A0">
<geo:lat>34.1000</geo:lat>
<geo:long>118.3333</geo:long>
</rdf:Description>
</rdf:RDF>

JSON-LD Serialization
{
"@context":{
"foaf": "http://xmlns.com/foaf/0.1/",
"child": {
"@id": "http://www.example.org/isFatherOf",
"@type": "@id"
},
"name": "foaf:name",
"location": "foaf:based_near",
"geo": "http://www.w3.org/2003/01/geo/wgs84_pos#",
"lat": "geo:lat",
"long": "geo:long"
},
"@id": "http://www.example.org/bradPitt",
"child": "http://www.example.org/maddoxJoliePitt",
"name": "Brad Pitt",
"location": {
"lat": "34.1000",
"long": "118.3333"
}
}

RDFa Serialization
<html>
<head>
…
</head>
<body vocab="http://xmlns.com/foaf/0.1/">
<div resource="http://www.example.org/bradPitt">
<p>Famous American actor <span property="name">Brad Pitt</span>
eldest son is
<a property="http://www.example.org/isFatheOf"
href="http://www.example.org/maddoxJoliePitt">Maddox Jolie-Pitt</a>.
</p>
</div>
</body>
</html>

The RDF Schema (1)
RDF
◦ A graph model
◦ Provides basic constructs for defining a graph structure
RDFS
◦ A semantic extension of RDF
◦ Provides mechanisms for assigning meaning to the RDF nodes and
edges
◦ RDF Schema is not to RDF what XML Schema is to XML!

The RDF Schema (2)
Classes
◦ A definition of groups of resources
◦ Members of a class are called instances of this class
Class hierarchy
Property hierarchy
Property domain and range
RDFS specification also refers to the RDF namespace
Prefix Namespace
rdfs http://www.w3.org/2000/01/rdf-schema#
rdf http://www.w3.org/1999/02/22-rdf-syntax-ns#

The RDF Schema (3)
A Class named Actor
◦ Where ex: http://www.example.org/
Class membership example
Containment relationship
Using reasoning, we can infer
◦ In practice, we could store inferred triples
ex:Actor rdf:type rdfs:Class.
ex:bradPitt rdf:type ex:Actor.
ex:Actor rdfs:subClassOf ex:MovieStaff.
ex:bradPitt rdf:type ex:MovieStaff.

The RDF Schema (4)
No restrictions on the form of the class hierarchy that can be defined
in an ontology
◦ No strict tree-like hierarchy as regular taxonomies
◦ More complex graph-like structures are allowed
◦ Classes may have more than one superclass
Allows hierarchies of properties to be defined
◦ Just like class hierarchies
◦ Introduced via the rdfs:subPropertyOf property
ex:participatesIn rdf:type rdf:Property.
ex:starsIn rdf:type rdf:Property.
ex:starsIn rdfs:subPropertyOf ex:participatesIn.

The RDF Schema (5)
◦ Inferred triple:
Define the classes to which RDF properties can be applied
◦ rdfs:domain and rdfs:range
Then, the assertion
Entails
ex:bradPitt ex:starsIn ex:worldWarZ.
ex:bradPitt ex:participatesIn ex:worldWarZ.
ex:georgeClooney ex:starsIn ex:idesOfMarch.
ex:georgeClooney rdf:type ex:Actor.
ex:starsIn rdfs:domain ex:Actor.

The RDF Schema (6)
Similarly
Produces
Domain and range axioms
◦ Are not constraints that data need to follow
◦ Function as rules that lead to the production of new triples and knowledge
ex:starsIn rdfs:range ex:Movie.
ex:worldWarZ rdf:type ex:Movie.
ex:starsIn rdfs:range ex:Movie.

The RDF Schema (7)
Example
◦ Infers:
An individual entity can be a member of both ex:Actor and ex:Movie
classes
Such restrictions are met in more expressive ontology languages,
such as OWL
ex:georgeClooney ex:starsIn ex:bradPitt.
ex:bradPitt rdf:type ex:Movie.

Reification (1)
Statements about other RDF statements
Ability to treat an RDF statement as an RDF resource
◦ Make assertions about that statement
Becomes:
<http://www.example.org/person/1> foaf:name "Brad Pitt " .
<http://www.example.org/statement/5> a rdf:Statement;
rdf:subject <http://www.example.org/person/1>;
rdf:predicate foaf:name;
rdf:object "Brad Pitt".

Reification (2)
Useful when referring to an RDF triple in order to
◦ Describe properties that apply to it
◦ e.g. provenance or trust
◦ E.g. assign a trust level of 0.8
<http://www.example.org/statement/5> ex:hasTrust "0.8"^^xsd:float.

Classes (1)
rdfs:Resource
◦ All things described by RDF are instances of the class rdfs:Resource
◦ All classes are subclasses of this class, which is an instance of
rdfs:Class
rdfs:Literal
◦ The class of all literals
◦ An instance of rdfs:Class
◦ Literals are represented as strings but they can be of any XSD
datatype

Classes (2)
rdfs:langString
◦ The class of language-tagged string values
◦ It is a subclass of rdfs:Literal and an instance of rdfs:Datatype
◦ Example: "foo"@en
rdfs:Class
◦ The class of all classes, i.e. the class of all resources that are RDF
classes

Classes (3)
rdfs:Datatype
◦ The class of all the data types
◦ An instance but also a subclass of rdfs:Class
◦ Every instance of rdfs:Datatype is also a subclass of rdfs:Literal
rdf:HTML
◦ The class of HTML literal values
◦ An instance of rdfs:Datatype
◦ A subclass of rdfs:Literal

Classes (4)
rdf:XMLLiteral
◦ The class of XML literal values
◦ An instance of rdfs:Datatype
◦ A subclass of rdfs:Literal
rdf:Property
◦ The class of all RDF properties

Properties (1)
rdfs:domain
◦ Declares the domain of a property P
◦ The class of all the resources that can appear as S in a triple (S, P, O)
rdfs:range
◦ Declares the range of a property P
◦ The class of all the resources that can appear as O in a triple (S, P, O)

Properties (2)
rdf:type
◦ A property that is used to state that a resource is an instance of a
class
rdfs:label
◦ A property that provides a human-readable version of a resource’s
name

Properties (3)
rdfs:comment
◦ A property that provides a human-readable description of a
resource
rdfs:subClassOf
◦ Corresponds a class to one of its superclasses
◦ A class can have more than one superclasses
rdfs:subPropertyOf
◦ Corresponds a property to one of its superproperties
◦ A property can have more than one superproperties

Container Classes and Properties (1)
rdfs:Container
◦ Superclass of all classes that can contain instances such as rdf:Bag, rdf:Seq and
rdf:Alt
rdfs:member
◦ Superproperty to all the properties that declare that a resource belongs to a
class that can contain instances (container)
rdfs:ContainerMembershipProperty
◦ A property that is used in declaring that a resource is a member of a container
◦ Every instance is a subproperty of rdfs:member

Container Classes and Properties (2)
rdf:Bag
◦ The class of unordered containers
◦ A subclass of rdfs:Container
rdf:Seq
◦ The class of ordered containers
rdf:Alt
◦ The class of containers of alternatives

Collections (1)
rdf:List
◦ The class of RDF Lists
◦ Can be used to build descriptions of lists and other list-like
structures
rdf:nil
◦ An instance of rdf:List that is an empty rdf:List

Collections (2)
rdf:first
◦ The first item in the subject RDF list
◦ An instance of rdf:Property
rdf:rest
◦ The rest of the subject RDF list after the first item
rdf:_1, rdf:_2, rdf:_3, etc.
◦ A sub-property of rdfs:member
◦ An instance of the class rdfs:ContainerMembershipProperty

Reification (1)
rdf:Statement
◦ The class of RDF statements
rdf:subject
◦ The subject of the subject RDF statement

Reification (2)
rdf:predicate
◦ The predicate of the subject RDF statement
rdf:object
◦ The object of the subject RDF statement

Utility Properties
rdf:value
◦ Idiomatic property used for describing structured values
rdfs:seeAlso
◦ A property that indicates a resource that might provide additional information
about the subject resource
rdfs:isDefinedBy
◦ A property that indicates a resource defining the subject resource
◦ The defining resource may be an RDF vocabulary in which the subject
resource is described

Outline
Introduction
RDF and RDF Schema
Description Logics
Other technologies
Ontologies
Datasets

Ontologies Based on Description Logics
Language roots are in Description Logics (DL)
DAML, OIL, and DAML + OIL
OWL and OWL 2 latest results in this direction
Formal semantics
Syntactically compatible to the RDF serializations
◦ Ontologies in OWL can be queried in the same approach as in RDF
graphs

Description Logics (1)
Offers the language for the description and manipulation of
independent individuals, roles, and concepts
There can be many DL languages
◦ Describe the world using formulas
Formulas constructed using
◦ Sets of concepts, roles, individuals
◦ Constructors, e.g. intersection ∧, union ∨, exists ∃, for each ∀

Variables in DL can represent arbitrary world objects
E.g. a DL formula can declare that a number x, greater than
zero exists:
◦ ∃x:greaterThan(x; 0)
Adding more constructors to the basic DL language increases
expressiveness
◦ Possible to describe more complex concepts

E.g. concept conjunction (C ∪ D)
◦ Parent = Father ∪ Mother
Expressiveness used in describing the world varies according
to the subset of the language that is used
Differentiation affects
◦ World description capabilities
◦ Behavior and performance of processing algorithms

The Web Ontology Language (1)
OWL language is directly related to DL
◦ Different “flavors” correspond to different DL language subsets
Successor of DAML+OIL
◦ Creation of the OWL language officially begins with the initiation of the DAML
project
◦ DAML, combined to OIL led to the creation of DAML + OIL
◦ An extension of RDFS
◦ OWL is the successor of DAML + OIL
W3C recommendation, currently in version 2

Designed in order to allow applications to process the
information content itself instead of simply presenting the
information
The goal is to provide a schema that will be compatible both
to the Semantic Web and the World Wide Web architecture
Makes information more machine- and human- processable

First version comprised three flavors
◦ Lite, DL, Full
OWL Lite
◦ Designed keeping in mind that it had to resemble RDFS
OWL DL
◦ Guaranteed that all reasoning procedures are finite and return a result
OWL Full
◦ The whole wealth and expressiveness of the language
◦ Reasoning is not guaranteed to be finite
◦ Even for small declaration sets

An OWL ontology may also comprise declarations from RDF
and RDFS
◦ E.g. rdfs:subClassOf, rdfs:range, rdf:resource
OWL uses them and relies on them in order to model its
concepts
rdfs:Resource
owl:Class
rdfs:Class rdf:Property
owl:DataProperty owl:ObjectProperty owl:AnnotationProperty

OWL 2
Very similar overall structure to the first version of OWL
Backwards compatible
◦ Every OWL 1 ontology is also an OWL 2 ontology

OWL 2 Additional Features (1)
Additional property and qualified cardinality constructors
◦ Minimum, maximum or exact qualified cardinality restrictions,
object and data properties
◦ E.g. ObjectMinCardinality, DataMaxCardinality
Property chains
◦ Properties as a composition of other properties
◦ E.g. ObjectPropertyChain in SubObjectPropertyOf

Extended datatype support
◦ Support the definition of subsets of datatypes (e.g. integers and
strings)
◦ E.g. state that every person has an age, which is of type integer, and restrict
the range of that datatype value
Simple metamodeling
◦ Relaxed separation between the names of, e.g. classes and
individuals
◦ Allow different uses of the same term

Extended annotations
◦ Allow annotations of axioms
◦ E.g. who asserted an axiom or when
Extra syntactic sugar
◦ Easier to write common patterns
◦ Without changing language expressiveness, semantics, or complexity

OWL 2 Profiles (1)
OWL 2 EL
◦ Polynomial time algorithms for all standard reasoning tasks
◦ Suitable for very large ontologies
◦ Trades expressive power for performance guarantees

OWL 2 Profiles (2)
OWL 2 QL
◦ Conjunctive queries are answered in LOGSPACE using standard
relational database technology
◦ Suitable for
◦ Relatively lightweight ontologies with large numbers of individuals
◦ Useful or necessary to access the data via relational queries (e.g. SQL)

OWL 2 Profiles (3)
OWL 2 RL
◦ Enables polynomial time reasoning algorithms using rule-extended
database technologies operating directly on RDF triples
◦ Suitable for
◦ Relatively lightweight ontologies with large numbers of individuals
◦ Necessary to operate directly on data in the form of RDF triples

Manchester OWL syntax
A user-friendly syntax for OWL 2 descriptions
Class: VegetarianPizza
EquivalentTo:
Pizza and
not (hasTopping some FishTopping) and
not (hasTopping some MeatTopping)
DisjointWith:
NonVegetarianPizza

Outline
Introduction
RDF and RDF Schema
Description Logics
Other technologies
Ontologies
Datasets

Querying the Semantic Web with SPARQL
SPARQL
◦ Is for RDF what SQL is for relational databases
SPARQL family of recommendations
◦ SPARQL Update
◦ SPARQL 1.1 Protocol
◦ Graph Store HTTP Protocol
◦ SPARQL 1.1 entailment regimes
◦ Specifications about the serialization of query results
◦ In JSON, CSV and TSV, and XML

SELECT Queries (1)
Return a set of bindings of variables to RDF terms
Binding
◦ A mapping from a variable to a URI or an RDF literal
Triple pattern
◦ Resembles an RDF triple
◦ May also contain variables in one or more positions

SELECT Queries (2)
An RDF triple matches a triple pattern
◦ There is an appropriate substitution of variables with RDF terms
that makes the triple pattern and the RDF triple equivalent
◦ Example:
matches
ex:bradPitt rdf:type ex:Actor.
?x rdf:type ex:Actor

SELECT Queries (3)
A set of triple patterns build a basic graph pattern
◦ The heart of a SPARQL SELECT query
SELECT ?x ?date
WHERE {
?x rdf:type ex:Actor .
?x ex:bornIn ?date
}

SELECT Queries (4)
Graph example
SPARQL query solution
?x ?date
ex:bradPitt "1963"^^xsd:gYear
ex:angelinaJolie "1975"^^xsd:gYear
ex:bradPitt rdf:type ex:Actor;
ex:bornIn "1963"^^xsd:gYear.
ex:angelinaJolie rdf:type ex:Actor;
ex:jenniferAniston rdf:type ex:Actor.

The OPTIONAL keyword
Resource ex:jenniferAniston not included in the results
◦ No RDF triple matches the second triple pattern
Use of the OPTIONAL keyword
◦ Retrieves all actors and actresses and get their birthdate only if it is specified
in the RDF graph
?x ?date
ex:bradPitt "1963"^^xsd:gYear
ex:angelinaJolie "1975"^^xsd:gYear
ex:jenniferAniston
SELECT ?x ?date
WHERE {
OPTIONAL {?x ex:bornIn ?date}
}

The UNION Keyword
Specify alternative graph patterns
Search for resources that are of type ex:Actor or ex:Director,
including cases where a resource may belong to both
classes
SELECT ?x
WHERE {
{?x rdf:type ex:Actor}
UNION
{?x rdf:type ex:Director}
}

The FILTER Keyword
Set a condition that limits the result of a SPARQL query
Return all actors that were known to be born before 1975
and their corresponding birth date
SELECT ?x ?date
WHERE{
?x ex:bornIn ?date .
FILTER(?date < "1975"^^xsd:gYear)
}

ORDER BY and LIMIT Keywords
SELECT ?x ?date
WHERE {
?x ex:bornIn ?date .
}
ORDER BY DESC(?date)
LIMIT 10

CONSTRUCT Queries
Generate an RDF graph based on a graph template using the
solutions of a SELECT query
CONSTRUCT {?x rdf:type ex:HighlyPaidActor}
WHERE {
?x ex:hasSalary ?salary .
FILTER (?salary > 1000000)
}

ASK Queries
Respond with a boolean
Whether a graph pattern is satisfied by a subgraph of the
considered RDF graph
Return true if the RDF graph contains an actor named "Cate
Blanchett"
ASK {
?x foaf:name "Cate Blanchett" .
}

DESCRIBE Queries
Return a set of RDF statements that contain data about a
given resource
Exact structure of the returned RDF graph depends on the
specific SPARQL engine
DESCRIBE <http://www.example.org/bradPitt> .

Outline
Introduction
RDF and RDF Schema
Description Logics
Other technologies
Ontologies
Datasets

Mapping Relational Data to RDF (1)
Lack of RDF data volumes → Slow uptake of the Semantic Web vision
Large part of available data is stored in relational databases
Several methods and tools were developed for the translation of
relational data to RDF
◦ Each one with its own set of features and, unfortunately, its own mapping
language
The need for the reuse of RDB-to-RDF mappings led to the creation
of R2RML
◦ A standard formalism by W3C for expressing such mappings

Mapping Relational Data to RDF (2)
R2RML: RDB to RDF Mapping Language
◦ Provides a vocabulary for the definition of RDF views over
relational schemas
◦ Is database vendor-agnostic
An R2RML mapping is an RDF graph: an R2RML mapping
graph
◦ In Turtle syntax

R2RML Overview
More features
◦ Organization of generated triples in named graphs
◦ Definition of blank nodes
◦ Specification of a generated literal’s language
RefObjectMap
TriplesMap
PredicateObjectMap
SubjectMap
Generated Triples
Generated Output Dataset
GraphMap
Join
ObjectMap
PredicateMap
LogicalTable

R2RML (1)
An R2RML mapping
◦ Associates relational views with functions that generate RDF terms
Logical tables
◦ Relations or (custom) views defined in the relational schema
Term maps
◦ RDF generating functions
◦ Distinguished according to the position of the RDF term in the
generated triple
◦ Subject, predicate, and object maps

R2RML (2)
Triples maps
◦ Functions that map relational data to a set of RDF triples
◦ Groups of term maps
◦ R2RML mappings contain one or more triples maps
Graph maps
◦ Generated RDF triples can be organized into named graphs

R2RML Example (1)
A relational instance
FILM
ID Title Year Director
1 The Hunger Games 2012 1
ACTOR
ID Name BirthYear BirthLocation
1 Jennifer Lawrence 1990 Louisville, KY
2 Josh Hutcherson 1992 Union, KY
FILM2ACTOR
FilmID ActorID
1 1
1 2
DIRECTOR
ID Name BirthYear
1 Gary Ross 1956

R2RML Example (2)
An R2RML
triples map
@prefix rr: <http://www.w3.org/ns/r2rml#>.
@prefix ex: <http://www.example.org/>.
@prefix dc: <http://purl.org/dc/terms/>.
<#TriplesMap1>
rr:logicalTable [ rr:tableName "FILM" ];
rr:subjectMap [
rr:template "http://data.example.org/film/{ID}";
rr:class ex:Movie;
];
rr:predicateObjectMap [
rr:predicate dc:title;
rr:objectMap [ rr:column "Title" ];
];
rr:predicate ex:releasedIn;
rr:objectMap [ rr:column "Year";
rr:datatype xsd:gYear;];
].

R2RML Example (3)
The result (in Turtle)
Note the reuse of terms from external ontologies
◦ E.g. dc:title from Dublin Core
<http://data.example.org/film/1> a ex:Movie;
dc:title "The Hunger Games";
ex:releasedIn "2012"^^xsd:gYear.

R2RML Example (4)
The R2RML mapping graph of the example
◦ Specifies the generation of a set of RDF triples for every row of the
FILM relation
◦ Is the simplest possible
◦ Contains only one triples map
Every triples map must have exactly
◦ One logical table
◦ One subject map
◦ One or more predicate-object maps

R2RML Example (5)
A logical table represents an SQL result set
◦ Each row gives rise to an RDF triple
◦ Or quad in the general case, when named graphs are used
A subject map is simply a term map that produces the
subject of an RDF triple

R2RML Example (6)
Term maps
◦ Template-valued (rr:template)
◦ The subject map of the example
◦ Constant-valued (rr:predicate)
◦ The predicate map of the example
◦ Column-valued (rr:column)
◦ The object map of the example
Every generated resource will be an instance of ex:Movie
◦ Use of rr:class

R2RML Example (7)
Subject maps
◦ Generate subjects
Predicate-object maps
◦ Generate pairs of predicates and objects
◦ Contain at least one predicate map
◦ Contain at least one object map
Typed Literals
◦ Use of rr:datatype
◦ The second object map of the example specifies that the datatype of the RDF
literal that will be generated will be xsd:gYear

Foreign Keys (1)
Add another predicate-
object map that operates
on the DIRECTOR relation
Specify the generation of
three RDF triples per row
of the DIRECTOR relation
@prefix foaf: <http://xmlns.com/foaf/0.1/>.
<#TriplesMap2>
rr:logicalTable [ rr:tableName "DIRECTOR" ];
rr:subjectMap [
rr:template "http://data.example.org/director/{ID}";
rr:class ex:Director;
];
rr:predicate foaf:name;
rr:objectMap [ rr:column "Name" ];
];
rr:predicate ex:bornIn;
rr:objectMap [ rr:column "BirthYear";
].

Foreign Keys (2)
Add another predicate-object
map, the referencing object
map
Link entities described in
separate tables
Exploit the foreign key
relationship of FILM.Director
and DIRECTOR.ID
<#TriplesMap1>
rr:logicalTable [ rr:tableName "FILM" ];
rr:subjectMap [
rr:template "http://data.example.org/film/{ID}";
rr:class ex:Movie;
];
rr:predicate dc:title;
rr:objectMap [ rr:column "Title" ];
];
rr:predicate ex:releasedIn;
rr:objectMap [ rr:column "Year";
rr:predicateObjectMap[
rr:predicate ex:directedBy;
rr:objectMap[
rr:parentTriplesMap <#TriplesMap2>;
rr:joinCondition[
rr:child "Director";
rr:parent "ID";
];
];
].

Foreign Keys (3)
TriplesMap1 now contains a referencing object map
◦ Responsible for the generation of the object of a triple
◦ Referencing object maps are a special case of object maps
◦ Follows the generation rules of the subject map of another triples
map
◦ Called parent triples map
◦ Join conditions are also specified in order to select the appropriate
row of the logical table of the parent triples map

Foreign Keys (4)
Not necessary for this foreign key relationship to be explicitly defined
as a constraint in the relational schema
R2RML is flexible enough to allow the linkage of any column set
among different relations
TriplesMap1 result
TriplesMap2 result
<http://data.example.org/film/1> ex:directedBy <http://data.example.org/director/1>.
<http://data.example.org/director/1> a ex:Director;
foaf:name "Gary Ross";

Custom Views (1)
R2RML view defined via the
rr:sqlQuery property
Used just as the native logical
tables
Could also have been defined as
an SQL view in the underlying
database system
◦ Not always feasible/desirable
Produced result set must not
contain columns with the same
name
@prefix dc: <http://purl.org/dc/terms/>.
<#TriplesMap3>
rr:logicalTable [ rr:sqlQuery """
SELECT ACTOR.ID AS ActorId, ACTOR.Name AS
ActorName, ACTOR.BirthYear AS ActorBirth, FILM.ID AS FilmId FROM
ACTOR, FILM, FILM2ACTOR WHERE FILM.ID=FILM2ACTOR.FilmID AND
ACTOR.ID=FILM2ACTOR.ActorID; """];
rr:subjectMap [
rr:template "http://data.example.org/actor/{ActorId}";
rr:class ex:Actor;
];
rr:predicate foaf:name;
rr:objectMap [ rr:column "ActorName" ];
];
rr:predicate ex:bornIn;
rr:objectMap [ rr:column "ActorBirth";
].
rr:predicate ex:starsIn;
rr:objectMap [ rr:template
"http://data.example.org/film/{ID}";
];
].

Custom Views (2)
TriplesMap3 generates the following triples
<http://data.example.org/actor/1> a ex:Actor;
foaf:name "Jennifer Lawrence";
ex:bornIn "1990"^^xsd:gYear;
ex:starsIn<http://data.example.org/film/1>.
<http://data.example.org/actor/2> a ex:Actor;
foaf:name "Josh Hutcherson";
ex:bornIn "1992"^^xsd:gYear;
ex:starsIn <http://data.example.org/film/1>.

Direct Mapping (1)
A default strategy for translating a relational instance to an
RDF graph
A trivial transformation strategy
A recommendation by W3C since 2012
Defines a standard URI generation policy for all kinds of
relations

Direct Mapping (2)
Maps every relation to a new ontology class and every relation
row to an instance of the respective class
The generated RDF graph essentially mirrors the relational
structure
Can be useful as a starting point for mapping authors who can
then augment and customize it accordingly using R2RML

Outline
Introduction
RDF and RDF Schema
Description Logics
Other technologies
Ontologies
Datasets

POWDER – Protocol for Web Description Resources (1)
XML-based protocol
Allows provision of information about RDF resources
A POWDER document contains
◦ Its own provenance information
◦ Information about its creator, when it was created, etc.
◦ A list of description resources

POWDER – Protocol for Web Description Resources (2)
Description resource
◦ Contains a set of property-value pairs in RDF
◦ For a given resource or group of resources
Allows for quick annotation of large amounts of content with
metadata
Ideal for scenarios involving personalized delivery of content,
trust control and semantic annotation

RIF – Rule Interchange Format
An extensible framework for the representation of rule dialects
A W3C recommendation
A family of specifications
◦ Three rule dialects
◦ RIF-BLD (Basic Logic Dialect)
◦ RIF-Core and
◦ RIF-PRD (Production Rule Dialect)
◦ RIF-FLD (Framework for Logic Dialects)
◦ Allows the definition of other rule dialects
◦ Specifies the interface of rules expressed in a logic-based RIF dialect with RDF and OWL

SPIN – SPARQL Inferencing Notation
RIF uptake not as massive as expected
◦ Complexity, lack of supporting tools
SPIN
◦ SPARQL-based
◦ Production rules
Links RDFS and OWL classes with constraint checks and
inference rules
◦ Similar to class behavior in object-oriented languages

GRDDL – Gleaning Resource Descriptions from Dialects of Languages
Mechanisms that specify how to construct RDF
representations
◦ Can be applied to XHTML and XML documents
◦ Defines XSLT transformations that generate RDF/XML
Is the standard way of converting an XML document to RDF
Familiarity with XSLT required

LDP – Linked Data Platform
Best practices and guidelines for organizing and accessing LDP resources
◦ Via HTTP RESTful services
◦ Resources can be RDF or non-RDF
◦ Specifies the expected behavior of a Linked Data server when it receives HTTP requests
◦ Creation, deletion, update
◦ Defines the concept of a Linked Data Platform Container
◦ A collection of resources (usually homogeneous)
LDP specification expected to establish a standard behavior for Linked Data
systems
Example: Apache Marmotta

Outline
Introduction
RDF and RDF Schema
Description Logics
Other technologies
Ontologies
Datasets

Ontologies and Datasets (1)
Standards and technologies (e.g. RDF and OWL)
◦ Machine-consumable serializations
◦ Syntactical groundwork
◦ Assignment of meaning to resources and their relationships
Vocabularies
◦ (Re)used in order for the described information to be commonly,
unambiguously interpreted and understood
◦ Reused by various data producers
◦ Make data semantically interoperable

Ontologies and Datasets (2)
Datasets
◦ Collections of facts involving specific individual entities and
ontologies, which provide an abstract, high-level, terminological
description of a domain, containing axioms that hold for every
individual
A great number of ontologies exist nowadays
◦ Subject domains: e.g. geographical, business, life sciences,
literature, media

Ontology search engines and specialized directories
Schemapedia
Watson
Swoogle
Linked Open Vocabularies (LOV)
◦ The most accurate and comprehensive source of ontologies used
in the Linked Data cloud
◦ Vocabularies described by appropriate metadata and classified to
domain spaces
◦ Full-text search enabled at vocabulary and element level

DC – Dublin Core (1)
A minimal metadata element set
Mainly used for the description of web resources
DC vocabulary
◦ Available as an RDFS ontology
◦ Contains classes and properties
◦ E.g. agent, bibliographic resource, creator, title, publisher, description

DC – Dublin Core (2)
DC ontology partitioned in two sets
◦ Simple DC
◦ Contains 15 core properties
◦ Qualified DC
◦ Contains all the classes and the rest of the properties (specializations of
those in simple DC)

FOAF – Friend-Of-A-Friend
One of the first lightweight ontologies being developed
since the first years of the Semantic Web
Describes human social networks
Classes such as Person, Agent or Organization
Properties denoting personal details and relationships with
other persons and entities

SKOS – Simple Knowledge Organization System (1)
A fundamental vocabulary in the Linked Data ecosystem
A W3C recommendation since 2009
Contains terms for organizing knowledge
◦ In the form of taxonomies, thesauri, and concept hierarchies
Defines concept schemes as sets of concepts
◦ Concept schemes can relate to each other through
“broader/narrower than” or equivalence relationships

SKOS – Simple Knowledge Organization System (2)
Defined as an OWL Full ontology
◦ Has its own semantics
◦ Its own set of entailment rules distinct from those of RDFS and OWL
Widely used in the library and information science domain

VoID – Vocabulary of Interlinked Datasets
Metadata for RDF datasets
◦ General metadata
◦ E.g. Name, description, creator
◦ Information on the ways that this dataset can be accessed
◦ E.g. as a data dump, via a SPARQL endpoint
◦ Structural information
◦ E.g. the set of vocabularies that are used or the pattern of a typical URI
◦ Information on the links
◦ E.g. the number and target datasets
Aid users and machines to decide whether a dataset is suitable for
their needs

SIOC – Semantically-Interlinked Online Communities
An OWL ontology
Describes online communities
◦ E.g. forums, blogs, mailing lists
Main classes
◦ E.g. forum, post, event, group, user
Properties
◦ Attributes of those classes
◦ E.g. the topic or the number of views of a post

Good Relations
Describes online commercial offerings
◦ E.g. Product and business descriptions, pricing and delivery
methods
Great impact in real-life applications
◦ Adoption by several online retailers and search engines
Search engines able to interpret product metadata
expressed in the Good Relations vocabulary
◦ Offer results better tailored to the needs of end users

Outline
Introduction
RDF and RDF Schema
Description Logics
Other technologies
Ontologies
Datasets

Datasets
Several “thematic neighborhoods” in the Linked Data cloud
◦ E.g. government, media, geography, life sciences
Single-domain and cross-domain datasets

DBpedia (1)
The RDF version of Wikipedia
◦ Perhaps the most popular RDF dataset in the LOD cloud
◦ A large number of incoming links
A cross-domain dataset
◦ Assigns a URI to every resource described by a Wikipedia article
◦ Produces structured information by mining information from
Wikipedia infoboxes

DBpedia (2)
A multilingual dataset
◦ Transforms various language editions of Wikipedia
The DBpedia dataset offered as data dump and through a
SPARQL endpoint
A “live” version available, updated whenever a Wikipedia
page is updated

Freebase
Openly-licensed, structured dataset, also available as an RDF
graph
Tens of millions of concepts, types and properties
Can be edited by anyone
Gathers information from several structured sources and free
text
Offers an API for developers
Linked to DBpedia through incoming and outgoing links

GeoNames
An open geographical database
Millions of geographical places
Can be edited by anyone
Geographical information as the context of descriptions and
facts
Omnipresent in the Linked Data cloud
Several incoming links from other datasets

Lexvo
Central dataset for the so-called linguistic Linked Data Cloud
Provides identifiers for thousands of languages
URIs for terms in every language
Identifiers for language characters

Chapter 3
Deploying Linked Open Data
Methodologies and Software Tools

Outline
Introduction
Modeling Data
Software for Working with Linked Data
Software Tools for Storing and Processing Linked Data
Tools for Linking and Aligning Linked Data
Software Libraries for working with RDF

Introduction
Today’s Web: Anyone can say anything about any topic
◦ Information on the Web cannot always be trusted
Linked Open Data (LOD) approach
◦ Materializes the Semantic Web vision
◦ A focal point is provided for any given web resource
◦ Referencing (referring to)
◦ De-referencing (retrieving data about)

Not All Data Can Be Published Online
Data has to be
◦ Stand-alone
◦ Strictly separated from business logic, formatting, presentation processing
◦ Adequately described
◦ Use well-known vocabularies to describe it, or
◦ Provide de-referenceable URIs with vocabulary term definitions
◦ Linked to other datasets
◦ Accessed simply
◦ HTTP and RDF instead of Web APIs

Linked Data-driven Applications (1)
Content reuse
◦ E.g. BBC’s Music Store
◦ Uses DBpedia and MusicBrainz
Semantic tagging and rating
◦ E.g. Faviki
◦ Uses DBpedia

Integrated question-answering
◦ E.g. DBpedia mobile
◦ Indicate locations in the user’s vicinity
Event data management
◦ E.g. Virtuoso’s calendar module
◦ Can organize events, tasks, and notes

Linked Data-driven data webs are expected to evolve in
numerous domains
◦ E.g. Biology, software engineering
The bulk of Linked Data processing is not done online
Traditional applications use other technologies
◦ E.g. relational databases, spreadsheets, XML files
◦ Data must be transformed in order to be published on the web

The O in LOD: Open Data
Open ≠ Linked
◦ Open data is data that is publicly accessible via internet
◦ No physical or virtual barriers to accessing them
◦ Linked Data allows relationships to be expressed among these data
RDF is ideal for representing Linked Data
◦ This contributes to the misconception that LOD can only be published in RDF
Definition of openness by www.opendefinition.org
Open means anyone can freely access, use, modify, and share for
any purpose (subject, at most, to requirements that preserve
provenance and openness)

Why should anyone open their data?
Reluctance by data owners
◦ Fear of becoming useless by giving away their core value
In practice the opposite happens
◦ Allowing access to content leverages its value
◦ Added-value services and products by third parties and interested audience
◦ Discovery of mistakes and inconsistencies
◦ People can verify convent freshness, completeness, accuracy, integrity, overall value
◦ In specific domains, data have to be open for strategic reasons
◦ E.g. transparency in government data

Steps in Publishing Linked Open Data (1)
Data should be kept simple
◦ Start small and fast
◦ Not all data is required to be opened at once
◦ Start by opening up just one dataset, or part of a larger dataset
◦ Open up more datasets
◦ Experience and momentum may be gained
◦ Risk of unnecessary spending of resources
◦ Not every dataset is useful

Engage early and engage often
◦ Know your audience
◦ Take its feedback into account
◦ Ensure that next iteration of the service will be as relevant as it can be
◦ End users will not always be direct consumers of the data
◦ It is likely that intermediaries will come between data providers and end users
◦ E.g. an end user will not find use in an array of geographical coordinates but a company offering
maps will
◦ Engage with the intermediaries
◦ They will reuse and repurpose the data

Deal in advance with common fears and misunderstandings
◦ Opening data is not always looked upon favorably
◦ Especially in large institutions, it will entail a series of
consequences and, respectively, opposition
◦ Identify, explain, and deal with the most important fears and
probable misconceptions from an early stage

It is fine to charge for access to the data via an API
◦ As long as the data itself is provided in bulk for free
◦ Data can be considered as open
◦ The API is considered as an added-value service on top of the data
◦ Fees are charged for the use of the API, not of the data
◦ This opens business opportunities in the data-value chain around
open data

Data openness ≠ data freshness
◦ Opened data does not have to be a real-time snapshot of the
system data
◦ Consolidate data into bulks asynchronously
◦ E.g. every hour or every day
◦ You could offer bulk access to the data dump and access through an API to
the real-time data

Dataset Metadata (1)
Provenance
◦ Information about entities, activities and people involved in the
creation of a dataset, a piece of software, a tangible object, a thing
in general
◦ Can be used in order to assess the thing’s quality, reliability,
trustworthiness, etc.
◦ Two related recommendations by W3C
◦ The PROV Data Model, in OWL 2
◦ The PROV ontology

Description about the dataset
W3C recommendation
◦ DCAT
◦ Describes an RDF vocabulary
◦ Specifically designed to facilitate interoperability between data catalogs
published on the Web

Licensing
◦ A short description regarding the terms of use of the dataset
◦ E.g. for the Open Data Commons Attribution License
This {DATA(BASE)-NAME} is made available under the Open Data Commons
Attribution License: http://opendatacommons.org/licenses/by/{version }.—See
more at: http://opendatacommons.org/licenses/by/#sthash.9HadQzSW.dpuf

Bulk Access vs. API (1)
Offering bulk access is a requirement
◦ Offering an API is not

Bulk access
◦ Can be cheaper than providing an API
◦ Even an elementary API entails development and maintenance costs
◦ Allows building an API on top of the offered data
◦ Offering an API does not allow clients to retrieve the whole amount of data
◦ Guarantees full access to the data
◦ An API does not

API
◦ More suitable for large volumes of data
◦ No need to download the whole dataset when a small subset is needed

The 5-Star Deployment Scheme
★
Data is made available on the Web (whatever format) but with an open
license to be Open Data
★★
Available as machine-readable structured data: e.g. an Excel spreadsheet
instead of image scan of a table
★★★
As the 2-star approach, in a non-proprietary format:
e.g. CSV instead of Excel
★★★★
All the above plus the use of open standards from W3C (RDF and SPARQL) to
identify things, so that people can point at your stuff
★★★★★
All the above, plus: Links from the data to other people’s data in order to
provide context

Outline
Introduction
Modeling Data

The D in LOD: Modeling Content
Content has to comply with a specific model
◦ A model can be used
◦ As a mediator among multiple viewpoints
◦ As an interfacing mechanism between humans or computers
◦ To offer analytics and predictions
◦ Expressed in RDF(S), OWL
◦ Custom or reusing existing vocabularies
◦ Decide on the ontology that will serve as a model
◦ Among the first decisions when publishing a dataset as LOD
◦ Complexity of the model has to be taken into account, based on the desired
properties
◦ Decide whether RDFS or one of the OWL profiles (flavors) is needed

Reusing Existing Works (1)
Vocabularies and ontologies have existed long before the
emergence of the Web
◦ Widespread vocabularies and ontologies in several domains
encode the accumulated knowledge and experience
Highly probable that a vocabulary has already been created
in order to describe the involved concepts
◦ Any domain of interest

Increased interoperability
◦ Use of standards can help content aggregators to parse and
process the information
◦ Without much extra effort per data source
◦ E.g. an aggregator that parses and processes dates from several
sources
◦ More likely to support the standard date formats
◦ Less likely to convert the formatting from each source to a uniform syntax
◦ Much extra effort per data source
◦ E.g. DCMI Metadata Terms
◦ Field dcterms:created, value "2014-11-07"^^xsd:date

Credibility
◦ Shows that the published dataset
◦ Has been well thought of
◦ Is curated
◦ A state-of-the-art survey has been performed prior to publishing the data
Ease of use
◦ Reusing is easier than rethinking and implementing again or replicating
existing solutions
◦ Even more, when vocabularies are published by multidisciplinary consortia with potentially
more spherical view on the domain than yours

In conclusion
◦ Before adding terms in our vocabulary, make sure they do not
already exist
◦ In such case, reuse them by reference
◦ When we need to be more specific, we can create a subclass or a
subproperty of the existing
◦ New terms can be generated, when the existing ones do not
suffice

Semantic Web for Content Modeling
Powerful means for system description
◦ Concept hierarchy, property hierarchy, set of individuals, etc.
Beyond description
◦ Model checking
◦ Use of a reasoner assures creation of coherent, consistent models
◦ Semantic interoperability
◦ Inference
◦ Formally defined semantics
◦ Support of rules
◦ Support of logic programming

Assigning URIs to Entities
Descriptions can be provided for
◦ Things that exist online
◦ Items/persons/ideas/things (in general) that exist outside of the Web
Example: Two URIs to describe a company
◦ The company’s website
◦ A description of the company itself
◦ May well be in an RDF document
A strategy has to be devised in assigning URIs to entities
◦ No deterministic approaches

Assigning URIs to Entities: Challenges
Dealing with ungrounded data
Lack of reconciliation options
Lack of identifier scheme documentation
Proprietary identifier schemes
Multiple identifiers for the same concepts/entities
Inability to resolve identifiers
Fragile identifiers

Assigning URIs to Entities: Benefits
Semantic annotation
Data is discoverable and citable
The value of the data increases as the usage of its identifiers
increases

URI Design Patterns (1)
Conventions for how URIs will be assigned to resources
Also widely used in modern web frameworks
In general applicable to web applications
Can be combined
Can evolve and be extended over time
Their use is not restrictive
◦ Each dataset has its own characteristics
Some upfront thought about identifiers is always beneficial

Hierarchical URIs
◦ URIs assigned to a group of resources that form a natural hierarchy
◦ E.g. :collection/:item/:sub-collection/:item
Natural keys
◦ URIs created from data that already has unique identifiers
◦ E.g. identify books using their ISBN

Literal keys
◦ URIs created from existing, non-global identifiers
◦ E.g. the dc:identifier property of the described resource
Patterned URIs
◦ More predictable, human-readable URIs
◦ E.g. /books/12345
◦ /books is the base part of the URI indicating “the collection of books”
◦ 12345 is an identifier for an individual book

Proxy URIs
◦ Used in order to deal with the lack of standard identifiers for third-party
resources
◦ If for these resources, identifiers do exist, then these should be reused
◦ If not, use locally minted Proxy URIs
Rebased URIs
◦ URIs constructed based on other URIs
◦ E.g. URIs rewritten using regular expressions
◦ From http://graph1.example.org/document/1 to
http://graph2.example.org/document/1

Shared keys
◦ URIs specifically designed to simplify the linking task between datasets
◦ Achieved by a creating Patterned URIs while applying the Natural Keys pattern
◦ Public, standard identifiers are preferable to internal, system-specific
URL slugs
◦ URIs created from arbitrary text or keywords, following a certain algorithm
◦ E.g. lowercasing the text, removing special characters and replacing spaces
with a dash
◦ A URI for the name “Brad Pitt” could be http://www.example.org/brad-pitt

Assigning URIs to Entities
Desired functionality
◦ Semantic Web applications retrieve the RDF
description of things
◦ Web browsers are directed to the (HTML)
documents describing the same resource
Two categories of technical approaches
for providing URIs for dataset entities
◦ Hash URIs
◦ 303 URIs
Resource Identifier (URI)
ID
Semantic Web
applications
Web
browsers
RDF document URI HTML document URI

Hash URIs
URIs contain a fragment separated from the rest of the URI
using ‘#’
◦ E.g. URIs for the descriptions of two companies
◦ http://www.example.org/info#alpha
◦ http://www.example.org/info#beta
◦ The RDF document containing descriptions about both companies
◦ http://www.example.org/info
◦ The original URIs will be used in this RDF document to uniquely
identify the resources
◦ Companies Alpha, Beta and anything else

Hash URIs with Content Negotiation
Redirect either to the RDF or the
HTML representation
◦ Decision based on client preferences
and server configuration
Technically
◦ The Content-Location header should be
set to indicate where the hash URI
refers to
◦ Part of the RDF document (info.rdf)
◦ Part of the HTML document (info.html)
http://www.example.org/info#alpha
Thing
application/rdf+xml
wins
text/html
wins
Content-Location:
http://www.example.org/info.html
Content-Location:
http://www.example.org/info.rdf
Automatic truncation of fragment
http://www.example.org/info

Hash URIs without Content Negotiation
Can be implemented by simply uploading
static RDF files to a Web server
◦ No special server configuration is needed
◦ Not as technically challenging as the previous one
Popular for quick-and-dirty RDF publication
Major problem: clients will be obliged to load
(download) the whole RDF file
◦ Even if they are interested in only one of the
resources
http://www.example.org/info#alpha
ID
http://www.example.org/info
Automatic truncation of fragment

303 URIs (1)
Approach relies on the “303 See Other” HTTP status code
◦ Indication that the requested resource is not a regular Web
document
Regular HTTP response (200) cannot be returned
◦ Requested resource does not have a suitable representation
However, we still can retrieve description about this
resource
◦ Distinguishing between the real-world resource and its description
(representation) on the Web

303 URIs (2)
HTTP 303 is a redirect status code
◦ Server provides the location of a document that represents the
resource
E.g. companies Alpha and Beta can be described using the
following URIs
◦ http://www.example.org/id/alpha
◦ http://www.example.org/id/beta
Server can be configured to answer requests to these URIs
with a 303 (redirect) HTTP status code

303 URIs (3)
Location can contain an HTML, an RDF, or any alternative form, e.g.
◦ http://www.example.org/doc/alpha
◦ http://www.example.org/doc/beta
This setup allows to maintain bookmarkable, de-referenceable URIs
◦ For both the RDF and HTML views of the same resource
A very flexible approach
◦ Redirection target can be configured separately per resource
◦ There could be a document for each resource, or one (large) document with
descriptions of all the resources

303 URIs (4)
303 URI solution based on a
generic document URI
http://www.example.org/id/alpha
Thing
303 redirect
application/rdf+xml
wins
text/html
wins
Content-Location:
http://www.example.org/doc/alpha.html
Content-Location:
http://www.example.org/doc/alpha.rdf
Generic Document
content
negotiation
http://www.example.org/doc/alpha

303 URIs (5)
303 URI solution without the generic
document URI
http://www.example.org/id/alpha
Thing
303 redirect
with content
negotiation
application/rdf+xml
wins
text/html
wins
http://www.example.org/company/alpha
http://www.example.org/data/alpha

303 URIs (6)
Problems of the 303 approach
◦ Latency caused by client redirects
◦ A client looking up a set of terms may use many HTTP requests
◦ While everything that could be loaded in the first request is there and is
ready to be downloaded
◦ Clients of large datasets may be tempted to download the full data
via HTTP, using many requests
◦ In these cases, SPARQL endpoints or comparable services should be
provided in order to answer complex queries directly on the server

303 URIs (7)
303 and Hash approaches are not mutually exclusive
◦ Contrarily, combining them could be ideal
◦ Allow large datasets to be separated into multiple parts and have identifiers
for non-document resources

Outline
Introduction
Modeling Data
Software for Working with Linked Data

Software for Working with Linked Data (1)
Working on small datasets is a task that can be tackled by
manually authoring an ontology, however
Publishing LOD means the data has to be programmatically
manipulated
Many tools exist that facilitate the effort

Software for Working with Linked Data (2)
The most prominent tools listed next
◦ Ontology authoring environments
◦ Cleaning data
◦ Software tools and libraries for working with Linked Data
◦ No clear lines can be drawn among software categories
◦ E.g. graphical tools offering programmatic access, or software libraries
offering a GUI

Ontology Authoring (1)
Not a linear process
◦ It is not possible to line up the steps needed in order to complete
its authoring
An iterative procedure
◦ The core ontology structure can be enriched with more
specialized, peripheral concepts
◦ Further complicating concept relations
◦ The more the ontology authoring effort advances, the more complicated the
ontology becomes

Ontology Authoring (2)
Various approaches
◦ Start from the more general and continue with the more specific
concepts
◦ Reversely, write down the more specific concepts and group them
Can uncover existing problems
◦ E.g. concept clarification
◦ Understanding of the domain in order to create its model
◦ Probable reuse and connect to other ontologies

Ontology Editors (1)
Offer a graphical interface
◦ Through which the user can interact
◦ Textual representation of ontologies can be prohibitively obscure
Assure syntactic validity of the ontology
Consistency checks

Ontology Editors (2)
Freedom to define concepts and their relations
◦ Constrained to assure semantic consistency
Allow revisions
Several ontology editors have been built
◦ Only a few are practically used
◦ Among them the ones presented next

Protégé (1)
Open-source
Maintained by the Stanford Center for Biomedical
Informatics Research
Among the most long-lived, complete and capable solutions
for ontology authoring and managing
A rich set of plugins and capabilities

Protégé (2)
Customizable user interface
Multiple ontologies can be developed in a single frame workspace
Several Protégé frames roughly correspond to OWL components
◦ Classes
◦ Properties
◦ Object Properties, Data Properties, and Annotation Properties
◦ Individuals
A set of tools for visualization, querying, and refactoring

Protégé (3)
Reasoning support
◦ Connection to DL reasoners like HermiT (included) or Pellet
OWL 2 support
Allows SPARQL queries
WebProtégé
◦ A much younger offspring of the Desktop version
◦ Allows collaborative viewing and editing
◦ Less feature-rich, more buggy user interface

TopBraid Composer
RDF and OWL authoring and editing environment
Based on the Eclipse development platform
A series of adapters for the conversion of data to RDF
◦ E.g. from XML, spreadsheets and relational databases
Supports persistence of RDF graphs in external triple stores
Ability to define SPIN rules and constraints and associate them with OWL classes
Maestro, Commercial, Free edition
◦ Free edition offers merely a graphical interface for the definition of RDF graphs and OWL
ontologies and the execution of SPARQL queries on them

The NeOn Toolkit
Open-source ontology authoring environment
◦ Also based on the Eclipse platform
Mainly implemented in the course of the EC-funded NeOn project
◦ Main goal: the support for all tasks in the ontology engineering life-cycle
Contains a number of plugins
◦ Multi-user collaborative ontology development
◦ Ontology evolution through time
◦ Ontology annotation
◦ Querying and reasoning
◦ Mappings between relational databases and ontologies
◦ ODEMapster plugin

Platforms and Environments
Data that published as Linked Data is not always produced primarily
in this form
◦ Files in hard drives, relational databases, legacy systems etc.
Many options regarding how the information is to be transformed
into RDF
Many software tools and libraries available in the Linked Data
ecosystem
◦ E.g. for converting, cleaning up, storing, visualizing, linking etc.
◦ Creating Linked Data from relational databases is a special case, discussed in
detail in the next Chapter

Cleaning-Up Data: OpenRefine (1)
Data quality may be lower than expected
◦ In terms of homogeneity, completeness, validity, consistency, etc.
Prior processing has to take place before publishing
It is not enough to provide data as Linked Data
◦ Published data must meet certain quality standards

Initially developed as “Freebase Gridworks”, renamed “Google
Refine” in 2010, “OpenRefine” after its transition to a community-
supported project in 2012
Created specifically to help working with messy data
Used to improve data consistency and quality
Used in cases where the primary data source are files
◦ In tabular form (e.g. TSV, CSV, Excel spreadsheets) or
◦ Structured as XML, JSON, or even RDF.

Allows importing data into the tool and connect them to other
sources
It is a web application, intended to run locally, in order to allow
processing sensitive data
Cleaning data
◦ Removing duplicate records
◦ Separating multiple values that may reside in the same field
◦ Splitting multi-valued fields
◦ Identifying errors (isolated or systematic)
◦ Applying ad-hoc transformations using regular expressions

OpenRefine: The RDF Refine Extension (1)
Allows conversion from other sources to RDF
◦ RDF export
◦ RDF reconciliation
RDF export part
◦ Describe the shape of the generated RDF graph through a template
◦ Template uses values from the input spreadsheet
◦ User can specify the structure of an RDF graph
◦ The relationships that hold among resources
◦ The form of the URI scheme that will be followed, etc.

OpenRefine: The RDF Refine Extension (2)
RDF reconciliation part
◦ Offers a series of alternatives for discovering related Linked Data
entities
◦ A reconciliation service
◦ Allows reconciliation of resources
◦ Against an arbitrary SPARQL endpoint
◦ With or without full-text search functionality
◦ A predefined SPARQL query that contains the request label (i.e. the label of the resource to be
reconciled) is sent to a specific SPARQL endpoint
◦ Via the Sindice API
◦ A call to the Sindice API is directly made using the request label as input to the service

Outline
Introduction
Modeling Data

Tools for Storing and Processing Linked Data
Storing and processing solutions
◦ Usage not restricted to these capabilities
A mature ecosystem of technologies and solutions
◦ Cover practical problems such as programmatic access, storage,
visualization, querying via SPARQL endpoints, etc.

Sesame
An open source, fully extensible and configurable with respect to
storage mechanisms, Java framework for processing RDF data
Transaction support
RDF 1.1 support
Storing and querying APIs
A RESTful HTTP interface supporting SPARQL
The Storage And Inference Layer (Sail) API
◦ A low level system API for RDF stores and inferences, allowing for various
types of storage and inference to be used

OpenLink Virtuoso (1)
RDF data management and Linked Data server solution
◦ Also a web application/web services/relational database/file
server
Offers a free and a commercial edition
Implements a quad store
◦ (graph, subject, predicate, object)

OpenLink Virtuoso (2)
Graphs can be
◦ Directly uploaded to Virtuoso
◦ Transient (not materialized) RDF views on top of its relational
database backend
◦ Crawled from third party RDF (or non-RDF, using Sponger) sources
Offers several plugins
◦ Full-text search, faceted browsing, etc.

Apache Marmotta
The LDClient library
◦ A Linked Data Client
◦ A modular tool that can convert data from other formats into RDF
◦ Can be used by any Linked Data project, independent of the Apache
Marmotta platform
Can retrieve resources from remote data sources and map their data
to appropriate RDF structures
A number of different backends is included
◦ Provide access to online resources
◦ E.g. Freebase, Facebook graph API, RDFa-augmented HTML pages

Callimachus
An open source platform
◦ Also available in an enterprise closed-source edition
Development of web applications based on RDF and Linked Data
A Linked Data Management System
◦ Creating, managing, navigating and visualizing Linked Data through
appropriate front-end components
Relies on XHTML and RDFa templates
◦ Populated by the results of SPARQL queries executed against an RDF triple
store
◦ Constitute the human-readable web pages

Visualization software
Users may have limited or non-existent knowledge of Linked Data and
the related ecosystem
LodLive
◦ Provides a navigator that uses RDF resources, relying on SPARQL endpoints
CubeViz
◦ A faceted browser for statistical data
◦ Relies on the RDF Data Cube vocabulary for representing statistical data in RDF
Gephi
GraphViz
Open-source, generic graph visualization platforms

Apache Stanbol
A semantic content management system
Aims at extending traditional CMS's with semantic services
Reusable components
◦ Via a RESTful web service API that returns JSON, RDF and supports JSON-LD
Ontology manipulation
Content enhancement
◦ Semantic annotation
Reasoning
Persistence

Stardog
RDF database, geared towards scalability
Reasoning
OWL 2
SWRL
Implemented in Java
Exposes APIs for Jena and Sesame
Offers bindings for its HTTP protocol in numerous languages
◦ Javascript, .Net, Ruby, Clojure, Python
Commercial and free community edition

Outline
Introduction
Modeling Data

The L in LOD (1)
Web of Documents
◦ HTML Links
◦ Navigate among (HTML) pages
Web of (Linked) Data
◦ RDF links
◦ Navigate among (RDF) data
◦ Relationships between Web resources
◦ Triples (resource, property, resource)
◦ Main difference from simple hyperlinks: they possess some meaning

The L in LOD (2)
Links to external datasets of the LOD cloud
Integration of the new dataset in the Web of data
Without links, all published RDF datasets would essentially
be isolated islands in the “ocean” of Linked Data

The L in LOD (3)
Establishing links can be done
◦ Manually
◦ I.e. the knowledge engineer identifies the most appropriate datasets and
external resources
◦ More suitable for small and static datasets
◦ Semi-automatically
◦ Harness existing open-source tools developed for such a purpose
◦ More suitable for large datasets

Silk (1)
An open-source framework for the discovery of links among
RDF resources of different datasets
Available
◦ As a command line tool
◦ With a graphical user interface
◦ Cluster edition, for the discovery of links among large datasets

Silk (2)
Link specification language
◦ Details and criteria of the matching process, including
◦ The source and target RDF datasets
◦ The conditions that resources should fulfill in order to be interlinked
◦ The RDF predicate that will be used for the link
◦ The pre-matching transformation functions
◦ Similarity metrics to be applied

LIMES
A link discovery framework among RDF datasets
Extracts instances and properties from both source and
target datasets, stores them in a cache storage or memory
and computes the actual matches
◦ Based on restrictions specified in a configuration file
Offers a web interface for authoring the configuration file

Sindice
A service that can be used for the manual discovery of
related identifiers
An index of RDF datasets that have been crawled and/or
extracted from semantically marked up Web pages
Offers both free-text search and SPARQL query execution
functionalities
Exposes several APIs that enable the development of Linked
Data applications that can exploit Sindice’s crawled content

DBpedia Spotlight
Designed for annotating mentions of DBpedia resources in text
Provides an approach for linking information from unstructured
sources to the LOD cloud through DBpedia
Tool architecture comprises:
◦ A web application
◦ A web service
◦ An annotation and an indexing API in Java/Scala
◦ An evaluation module

Sameas.org
An online service
Retrieves related LOD entities from some of the most
popular datasets
Serves more than 150 million URIs
Provides a REST interface that retrieves related URIs for a
given input URI or label
Accepts URIs as inputs from the user and returns URIs that
may well be co-referent

Outline
Introduction
Modeling Data

Jena (1)
Open-source Java API
Allows building of Semantic Web and Linked Data applications
The most popular Java framework for ontology manipulation
First developed and brought to maturity by HP Labs
Now developed and maintained by the Apache Software Foundation
First version in 2000
Comprises a set of tools for programmatic access and management
of Semantic Web resources

Jena (2)
ARQ, a SPARQL implementation
Fuseki
◦ SPARQL Server, part of the Jena framework, that offers access to the data over
HTTP using RESTful services. Fuseki can be downloaded and extracted locally,
and run as a server offering a SPARQL endpoint plus some REST commands to
update the dataset.
OWL support
◦ Coverage of the OWL language
Inference API
◦ Using an internal rule engine, Jena can be used as a reasoner

Jena TDB (1)
High-performance triple store solution
Based on a custom implementation of threaded B+ Trees
◦ Only provides for fixed length key and fixed length value
◦ No use of the value part in triple indexes
Efficient storage and querying of large volumes of graphs
◦ Performs faster, scales much more than a relational database
backend
Stores the dataset in directory

Jena TDB (2)
A TDB instance consists of
◦ A Node table that stores the representation of RDF terms
◦ Triple and Quad indexes
◦ Triples are used for the default graph, quads for the named graphs
◦ The Prefixes table
◦ Provides support for Jena’s prefix mappings and serves mainly presentation
and serialization of triples issues, and does not take part in query processing

Apache Any23 (1)
A programming library
A web service
A command line tool
Ability to extract structured data in RDF from Web documents
Input formats
◦ RDF (RDF/XML, Turtle, Notation 3)
◦ RDFa, microformats (hCalendar, hCard, hListing, etc.)
◦ HTML 5 Microdata (such as schema.org)
◦ JSON-LD (Linked Data in JSON format)

Apache Any23 (2)
Support for content extraction following several vocabularies
◦ CSV
◦ Dublin Core Terms
◦ Description of a Career
◦ Description Of A Project
◦ Friend Of A Friend
◦ GeoNames
◦ ICAL
◦ Open Graph Protocol
◦ schema.org
◦ VCard, etc.

Redland
Support for programmatic management and storage of RDF graphs
A set of libraries written in C, including:
◦ Raptor
◦ Provides a set of parsers and serializers of RDF
◦ Including RDF/XML, Turtle, RDFa, N-Quads, etc.
◦ Rasqal
◦ Supports SPARQL query processing of RDF graphs
◦ Redland RDF Library
◦ Handles RDF manipulation and storage
Also allows function invocation through Perl, PHP, Ruby, Python, etc.

EasyRDF
A PHP library for the consumption and production of RDF
Offers parsers and serializers for most RDF serializations
Querying using SPARQL
Type mapping from RDF resources to PHP objects

RDFLib
A Python library for working with RDF
Serialization formats
Microformats
RDFa
OWL 2 RL
Using relational databases as a backend
Wrappers for remote SPARQL endpoints

The Ruby RDF Project
RDF Processing using the Ruby language
Reading/writing in different RDF formats
Microdata support
Querying using SPARQL
Using relational databases as a storage backend
Storage adaptors
◦ Sesame
◦ MongoDB

dotNetRDF (1)
Open-source library for RDF
Written in C#.Net, also offering ASP.NET integration
Developer API

dotNetRDF (2)
Supports
◦ SPARQL
◦ Reasoning
◦ Integration with third party triple stores
◦ Jena, Sesame, Stardog, Virtuoso, etc.
Suite of command line and graphical tools for
◦ Conversions between RDF formats
◦ Running a SPARQL server and submitting queries
◦ Managing any supported triple stores

Chapter 4
Creating Linked Data from
Relational Databases

Outline
Introduction
Motivation-Benefits
Classification of approaches
Creating ontology and triples from a relational database
Complete example
Future outlook

Introduction (1)
Relational databases vs. Semantic Web standards
◦ Active research topic since more than a decade ago
◦ Not just a theoretical exercise, but also practical value
◦ Bootstrap the Semantic Web with a sufficiently large mass of data
◦ Facilitate database integration
◦ Ontology-based data access
◦ Semantic annotation of dynamic Web pages

Introduction (2)
Database-to-ontology mapping
◦ The investigation of the similarities and differences among
relational databases and Semantic Web knowledge models
◦ Broad term encompassing several distinct problems
◦ Classification of approaches needed

Outline
Introduction
Motivation-Benefits
Complete example
Future outlook

Semantic Annotation of Dynamic Web Pages (1)
Goal of the Semantic Web: emergence of a Web of Data,
from the current Web of Documents
HTML documents are mainly for human consumption
How to achieve this?
◦ Add semantic information to HTML documents
◦ I.e. setup correspondences with terms from ontologies
RDFa
◦ Embedding references to ontology terms in XHTML tags, but…

What about dynamic documents?
◦ Content retrieved from relational databases
◦ The biggest part of the World Wide Web
◦ Aka. Deep Web
◦ CMS's, forums, wikis, etc.
◦ Manual annotation of every single dynamic web page is infeasible

Directly “annotate” the database schema!
◦ Establish correspondences between the elements of the database
schema and a suitable existing domain ontology
Use these correspondences to generate automatically
semantically annotated dynamic pages

Heterogeneous Database Integration (1)
Longstanding issue in database research
◦ Due to differences in:
◦ Software infrastructure
◦ Syntax
◦ Representation models
◦ Interpretation of the same data
Remains unresolved to a large degree

Typical database integration architecture
◦ One or more conceptual models for the description of the
contents of each source database
◦ Queries against a global conceptual schema
◦ Wrappers on top of every source database for the reformulation of
queries and data retrieval

Ontology-based database integration
◦ Ontologies instead of conceptual schemas
◦ Definition of correspondences between source databases and one
or more ontologies
◦ LAV, GAV or GLAV approach (target schema = ontology)
◦ Database term ↔ Query over the ontology (LAV)
◦ Ontology term ↔ Query over the database (GAV)
◦ Query over the database ↔ Query over the ontology (GLAV)
◦ Mappings between relational database schemas and ontologies
need to be discovered!

Ontology-Based Data Access (1)
Objective:
◦ Offer high-level services on top of an information system without
knowledge of the underlying database schema
Ontology as an intermediate layer between the end user
and the storage layer
◦ Ontology provides an abstraction of the database contents
◦ Users formulate queries using terms from the ontology

Ontology-Based Data Access (2)
Similar to a database integration architecture
◦ OBDA engine ≈ wrapper
◦ Transforms queries against the ontology to queries against the local data source
OBDA engine
◦ Performs query rewriting
◦ Uses mappings between a database and a relevant domain ontology
Advantages
◦ Semantic queries posed directly to the database
◦ No need to replicate database contents in RDF

Semantic Rewriting of SQL Queries
Objective:
◦ Reformulate an SQL query to another one that better captures the intention
of the user
Substitution of terms in the original SQL query with synonyms and
related terms from an ontology
Also related:
◦ Query relational data using external ontologies as context
◦ SQL queries with their WHERE conditions containing terms from an ontology
Feature implemented in some DBMSes
◦ E.g. OpenLink Virtuoso, Oracle

Mass Data Generation for the Semantic Web
Reasons for slow uptake of the Semantic Web
◦ Few successful paradigms of tools and “killer” applications
◦ Few data
◦ “Chicken-and-egg” problem
Relational databases hold the majority of data on the World Wide
Web
Automated extraction of RDB contents in RDF
Generation of a critical mass of Semantic Web data
Increased production of SW applications and tools anticipated

Ontology Learning (1)
Manual development of ontologies is difficult, time-
consuming and error-prone
Ontology learning
◦ Semi-automatic extraction of ontologies from free texts, semi-
structured documents, controlled vocabularies, thesauri etc.
◦ Relational databases can be sources of domain knowledge as well
◦ Information gathered from database schema, contents, queries
and stored procedures
◦ Supervision from domain expert is necessary

Ontology Learning (2)
Useful in domains where there is no suitable ontology
◦ Typical in the earlier Semantic Web years
Nowadays, ontology learning for the creation of a
“wrapping” ontology for an RDB in:
◦ OBDA
◦ Database integration

Intended Meaning of a Relational Schema (1)
Database schema design
◦ Conceptual model → relational model
◦ Subsequent changes often directly to the relational model
◦ Initial conceptual model lost
◦ Hard to re-engineer to another model (e.g. object-oriented)
Definition of correspondences between RDB and ontology
◦ Semantic grounding of the meaning of the former

Intended Meaning of a Relational Schema (2)
Facilitates:
◦ Database maintenance
◦ Integration with other data sources
◦ Mapping discovery between 2 or more database schemas
In the latter case, database-to-ontology mappings are used
as a reference point for the construction of inter-database
schema mappings

Database Integration with Other Data Sources
Mapping RDB to RDF enables integration with existing RDF content
◦ Content generated from either structured or unstructured sources
Linked Data paradigm
◦ Vocabulary reuse
◦ Inter-dataset links
◦ Identifier reuse
◦ Facilitates data source integration at global level
◦ Billions of RDF statements from several domains of interest
Integration of RDB content with Linked Data offers unlimited potential

Outline
Introduction
Motivation-Benefits
Complete example
Future outlook

Existing Classifications (1)
Several classification schemes proposed for database-to-
ontology mapping approaches
Classification criteria vs. descriptive measures
◦ Classification criteria
◦ Finite number of values
◦ Should separate approaches in non-overlapping sets
◦ Descriptive measures
◦ Can also be qualitative

Work Classification criteria Values Descriptive parameters
(Auer et al. 2009) a. Automation in the creation of mapping
b. Source of semantics considered
c. Access paradigm
d. Domain reliance
a. Automatic/Semi-automatic/Manual
b. Existing domain
ontologies/Database/Database and User
c. Extract-Transform-Load (ETL)/SPARQL/Linked
Data
d. General/Dependent
Mapping representation
language
(Barrasa-Rodriguez and
Gómez-Pérez 2006)
a. Existence of ontology
b. Architecture
c. Mapping exploitation
Yes (ontology reuse)/No (created ad-hoc)
Wrapper/Generic engine and declarative
definition
Massive upgrade (batch)/Query driven (on
demand)
-
(Ghawi and Cullot 2007) a. Existence of ontology
b. Complexity of mapping definition
c. Ontology population process
d. Automation in the creation of mapping
a. Yes/No
b. Complex/Direct
c. Massive dump/Query driven
d. Automatic/Semi-automatic/Manual
Automation in the instance
export process
(Hellmann et al. 2011)
- -
Data source, Data exposition, Data synchronization,
Mapping language, Vocabulary reuse, Mapping
automation,
Requirement of domain ontology, Existence of GUI

Work Classification criteria Values Descriptive parameters
(Konstantinou et al. 2008) a. Existence of ontology
b. Automation in the creation of mapping
c. Ontology development
a. Yes/No
b. Automatic/Semi-
automatic/Manual
c. Structure driven/Semantics
driven
Ontology language,
RDBMS supported, Semantic
query language,
Database components
mapped, Availability of
consistency checks, User
interaction
(Sahoo et al. 2009)
Same as in (Auer et al. 2009) with the addition of:
a. Query implementation
b. Data integration
a. SPARQL/SPARQLSQL
b. Yes/No
Mapping accessibility, Application
domain
(Sequeda et al. 2009) - - Correlation of primary and
foreign keys, OWL and
RDFS elements mapped
(Zhao and Chang 2007) a. Database schema analysis Yes/No Purpose, Input, Output,
Correlation analysis of
database schema elements,
Consideration of database
instance, application source
code and other sources

A Proposed Classification (1)
Relational DBs to Semantic Web
New ontologyExisting ontology
Database schema ontology Domain-specific ontology
No database reverse
engineering
Database reverse
engineering
• semantic annotation of dynamic web pages
• mass generation of SW data
• definition of meaning of relational schema
• heterogeneous database integration
• ontology based data access
• integration with other data sources
• semantic annotation of dynamic web pages
• integration with other data sources
• ontology learning

A Proposed Classification (2)
Total classification of all relevant solutions in mutually disjoint classes
Exceptions
◦ Customizable software tools with multiple possible workflows
◦ Each one belongs to multiple categories
Every class associated with a number of benefits/motivations
◦ Not significant correlation among taxonomy classes and motivations and benefits
◦ Categorization of approaches based on the nature of the mapping and the
techniques applied to establish the mapping
◦ Benefits state the applications of the already established mappings

Classification Criteria (1)
Existence of ontology
◦ Is an ontology required for the application of the approach?
◦ Yes
◦ Establishment of mappings between a given relational database and a given existing ontology
◦ Domain of ontology compatible with database domain
◦ Existing ontology selected by human user
◦ No
◦ Creation of a new ontology from a given relational database
◦ Useful when:
◦ An ontology for the domain covered by the database is not available yet
◦ The human user is not familiar with the domain of the database and relies on the mapping process to
discover the semantics of the database contents

Domain of the generated ontology
◦ What is the domain of the generated ontology?
◦ The relational model
◦ Generated ontology consists of concepts and relationships that reflect the
constructs of the relational model
◦ Mirrors the structure of the input relational database
◦ “Database schema ontology”
◦ Mainly automatic class of approaches
◦ Another domain
◦ Depending on the domain described by the contents of the input database

Database reverse engineering
◦ Are any database reverse engineering techniques applied?
◦ Yes
◦ Recover the initial conceptual schema from the relational schema
◦ Translate re-engineered schema to an ontology expressed in a target language
◦ No
◦ Few basic translation rules from the relational to the RDF model
◦ Reliance on the human expert for the definition of complex mappings and the
enrichment of the generated ontology

Classification criteria and descriptive features
Existence of ontology
• Yes
• No
Ontology domain
• Relational model
• Other
Application of database reverse
engineering
• Yes
• No
Automation level
• Automatic
• Semi-automatic
• Manual
Data accessibility
• ETL
• SPARQL
• Linked Data
Mapping Language
• SQL
• RDF/XML
• Custom language
Ontology language
• RDFS
• OWL dialect
Vocabulary reuse
• Yes
• No
Software availability
• Yes
• No
• Commercial
Graphical user interface
• Yes
• No
Purpose
• Mass generation of SW data
• Ontology learning
• Ontology based data access
• Database integration

Descriptive Features (1)
Level of Automation
◦ How much is the user involved in the mapping process?
◦ Automatic
◦ No input from human user
◦ Semi-automatic
◦ Some input from human user
◦ Sometimes necessary
◦ Sometimes optional (e.g. validation or enrichment of results)
◦ Manual
◦ Mapping defined entirely from human user
◦ Feature usually common among approaches of the same class

Data Accessibility
◦ The way the mapping result is accessed
◦ Aka. access paradigm / mapping implementation / data exposition
◦ ETL
◦ Result of the mapping process generated and stored as a whole in an
external storage medium (i.e. materialized)
◦ Aka. batch transformation / massive dump

Data Accessibility (cont’d)
◦ SPARQL
◦ Only a part of the mapping result is accessed
◦ No additional storage medium is required (i.e. no materialization)
◦ Rewriting of a SPARQL query to an SQL one
◦ SQL results transformed back to SPARQL results
◦ Aka. query-driven access
◦ Linked Data
◦ Mapping result published as Linked Data (i.e. all URIs use the HTTP scheme and,
when dereferenced, provide useful information for the resource they identify)

Data Synchronization
◦ Does the mapping result reflect the current database contents?
◦ Static
◦ Mapping executed only once
◦ Mapping result not tied with source database
◦ Dynamic
◦ Mapping executed on every incoming query
◦ Mapping result depends on current database state
◦ Strongly related to data accessibility, redundant feature
◦ ETL methods are static
◦ SPARQL (query-driven) and Linked Data methods are dynamic

Mapping language
◦ The language in which the mapping is represented
◦ Large variance of values: a lot of proprietary formats
◦ …until the standardization of R2RML
◦ Feature only applicable to methods that need to reuse the
mapping
◦ E.g. not applicable to ontology generation methods

Ontology language
◦ The language in which the involved ontology is expressed
◦ Either:
◦ The language of the ontology generated by the approach
◦ The language of the existing ontology required
◦ RDFS
◦ OWL (all flavours and dialects)

Vocabulary reuse
◦ Does the mapping support more than one existing ontologies?
◦ Yes
◦ Mainly manual approaches
◦ Human user free to reuse terms from existing ontologies
◦ Not obligatory to reuse terms
◦ No
◦ E.g. methods generating a new “database schema ontology”

Software availability
◦ Does the method have a free implementation?
◦ Theoretical methods
◦ Practical solutions
◦ Commercial software

Graphical User Interface
◦ Can the user interact with the system via a GUI?
◦ Feature applicable to approaches with an accessible software
implementation
◦ Guides user through steps of the mapping process
◦ Provides mapping suggestions
◦ Essential for inexperienced users / users not familiar with SW
technologies

Outline
Introduction
Motivation-Benefits
Complete example
Future outlook

Creating Ontology and Triples from a Relational Database (1)
Relational
database
Mapping
engine
Mappings
Rules
File storage
Persistent
storage
1
2
2 3
SQL
3

Creating Ontology and Triples from a Relational Database (2)
Generation of a new ontology
Population with RDF data originating from the database
Mapping engine
◦ Communicates with database
◦ Uses heuristic or manually defined rules
3 ways to access the generated RDF data
◦ ETL
◦ SPARQL
◦ Linked Data

The Basic Approach (1)
Method proposed by Tim Berners-Lee (1998)
Generic, applicable to every database
Automatic
“Table-to-class, column-to-predicate” method
A URI generation scheme also needed
◦ Should be reversible (i.e. recognize database element from URI)

Rules:
(a) Every relation R maps to an RDFS class C(R)
(b) Every tuple of a relation R maps to an RDF node of type C(R)
(c) Every attribute att of a relation maps to an RDF property P(att)
(d) For every tuple R[t], the value of an attribute att maps to a
value of the property P(att) for the node corresponding to the
tuple R[t]

db: database name
rel: relation name
attr: attribute name
pk: name of a primary key
pkval: value of primary key for given tuple
Database Element URI Template Example
Database {base_URI}/{db} http://www.example.org/company_db
Relation {base_URI}/{db}/{rel} http://www.example.org/company_db/emp
Attribute {base_URI}/{db}/{rel}#{attr} http://www.example.org/company_db/emp#name
Tuple {base_URI}/{db}/{rel}/{pk=pkval} http://www.example.org/company_db/emp/id=5
Typical URI generation scheme

Very crude export
Simple generated ontology
◦ No complex constructs
◦ Looks like a copy of the relational schema
New URI for every tuple
◦ Even when there is an existing one for an entity
All database values mapped to literals
◦ “Flat” RDF graph
Nevertheless, serves as foundation for several approaches

Creation and Population of a Domain Ontology (1)
“Database schema ontologies” are hardly useful for Linked Data publication
Domain-specific ontologies reflect the domain of the database
Expressiveness of generated ontology depends on the amount of domain
knowledge extracted from:
◦ Human user
◦ Relational instance
a) Approaches using database schema reverse engineering
b) Basic approach + enrichment from human user
More tools follow b)
◦ User has full control of the mapping

Automation level
◦ Depends on the involvement of the human user
Data accessibility
◦ SPARQL-based access more popular
Mapping language
◦ Needed to express complex correspondences between database and ontology
◦ Until R2RML, every tool used its own language
◦ Mapping lock-in, low interoperability

Ontology language
◦ RDFS, since majority of tools follows basic approach
Vocabulary reuse
◦ Possible when mappings are manually defined
◦ User should be familiar with SW vocabularies

Main goal
◦ Generate lightweight ontologies reusing existing terms
◦ Increased semantic interoperability
◦ Focus not on ontology expressiveness
Motivation
◦ Mass generation of RDF data from existing large quantities of
relational data
◦ Easier integration with other heterogeneous data

D2RQ / D2R Server (1)
One of the most popular tools in the field
Both automatic and user-assisted operation modes
◦ Automatic mode
◦ Automatic mapping generation
◦ Basic approach + rules for M:N relationships → RDFS ontology
◦ Semi-automatic mode
◦ User modifies automatic mapping
◦ Manual mode
◦ User builds mapping from scratch

D2RQ / D2R Server (2)
Custom mapping language
◦ Feature-rich
◦ URI generation mechanism
◦ Translation schemes for database values etc.
Both ETL and SPARQL-based access
Vocabulary reuse
◦ Refer to any ontology inside the mapping file

OpenLink Virtuoso Universal Server
Integration platform (both commercial and open-source versions)
RDF Views feature
◦ Similar functionality to D2RQ
Both automatic and manual modes
◦ Automatic mode relies on the basic approach
Virtuoso Meta-Schema language for the mapping definition
◦ Also very expressive
◦ One has to learn it in order to customize the mapping (same as in D2RQ)
ETL, SPARQL-based and Linked Data access

Triplify
RDF extraction tool from relational instances
Maps subsets of the database contents (i.e. SQL queries) to URIs of ontology
terms
◦ No need for users to learn a new mapping language
Mappings as configuration files
◦ Can reuse terms from existing vocabularies (manual editing)
ETL (static) and Linked Data (dynamic) access
Predefined mappings for schemas used by popular Web applications
Supports update logs for RDF resources
◦ Useful for crawling engines

Ultrawrap
Wraps a database as a SPARQL endpoint
Commercial tool
Supports creation of new domain ontology
◦ Set of advanced heuristic rules
SPARQL-based access
◦ SPARQL query refers to terms from new ontology
◦ Mappings expressed as views defined on the relational schema
◦ Rewriting to SQL queries referring to above views
Support for manual mappings that reuse terms from existing vocabularies

Oracle DBMS
RDF Views feature (similar to Virtuoso)
Query relational data as RDF
◦ No replication
◦ No physical storage for RDF graphs
Both automatic and manual mappings
◦ Automatic mode follows W3C’s Direct Mapping
Supports combination of virtual and materialized RDF data in the
same query

Mapping a Database to an Existing Ontology (1)
Domain ontologyRelational database
Mapping
engine
Mappings1
1
Schema
matching
algorithm
Mapping
execution
module
Mappings
2
RDF Graph

Existence of ontology is required
◦ Assumption: Ontology domain same as database domain
Discover mappings between a database and an ontology
◦ Schema matching algorithms
◦ Reverse engineering + linguistic similarity measures
◦ Reuse of such mappings in other applications (e.g. database integration)

Apply user-defined mappings to a database
◦ Mappings refer to one or more existing ontologies
◦ RDF graph contains instance data from the database
◦ Tools useful for Linked Data publication

Ontop (1)
Conversion of a relational instance to a SPARQL endpoint
User-defined mappings
Ontology-based data access (OBDA) framework
◦ Not just SPARQL
◦ RDFS and OWL 2 QL entailment regimes

Ontop (2)
No need to materialize inferences, calculated at query-time
SPARQL-to-SQL rewriting
◦ Datalog as intermediate representation language
◦ Several optimizations simplifying generated SQL queries
Plugin for ontology editor Protégé also available

R2O / ODEMapster / Morph
Declarative XML-based mapping language
Support for complex mappings
◦ Conditional mappings
◦ Definition of URI generation scheme
ODEMapster engine
◦ R2O mappings
◦ Materialized / query-driven access
Morph
◦ R2RML mappings
◦ SPARQL-based data access

R2RML Parser
Export of RDF graphs from a relational instance
R2RML mappings
Materialized RDF graph (ETL)
Supports faceted browsing of the generated RDF graph
Incremental dump feature
◦ Tackles the data synchronization issue
◦ Graph not generated from scratch
◦ Only the necessary updates are made to the extracted RDF graph

Outline
Introduction
Motivation-Benefits
Complete example
Future outlook

Linked Data in Scholarly/Cultural Heritage Domain (1)
Rich experience
Software systems that demonstrate flawless performance
High level of accuracy
Why evolve?
◦ Data and knowledge description
◦ New technologies entail new benefits
◦ Solutions have to remain competitive

Solutions by the LOD paradigm
◦ Integration
◦ Typically materialized using OAI-PMH that does not ease integration with
data from other domains
◦ Expressiveness in describing the information
◦ OAI-PMH allows for a tree structure that extends to a depth-level of two
◦ RDF allows for a graph-based description
◦ Query answering
◦ Querying graphs using graph patterns allows for much more complex queries

Benefits
◦ Query expressiveness
◦ Inherent semantics
◦ Integration with third party sources
Disadvantages
◦ Resources investment in creating and maintaining the data

More and more institutions open their data
◦ Biblioteca Nacional De España
◦ Deutsche National Bibliothek
◦ British Library

Is Linked Data the future?
◦ Content re-use
◦ Participation of individual collections
◦ Evolving global Linked Data cloud
◦ Users can discover new data sources following data-level links
◦ More complete answers can be delivered as new data sources
appear

Ontologies Related to Scholarly Information (1)
Good practice
◦ Reuse existing vocabularies/ontologies
◦ Easier for the outside world to integrate with already existing datasets and
services
◦ Several vocabularies have been proposed

Ontologies Related to Scholarly Information (2)
Title URL Namespace Namespace URL
The Bibliographic Ontology bibliontology.com bibo http://purl.org/ontology/bibo/
Creative Commons Rights Ontology creativecommons.org cc http://creativecommons.org/ns#
CiTo, the Citation Typing Ontology purl.org/spar/cito cito http://purl.org/spar/cito/
Legacy Dublin Core element set dublincore.org/documents/dces/ dc http://purl.org/dc/elements/1.1/
DCMI Metadata Terms dublincore.org/documents/dcmi-terms/ dcterms http://purl.org/dc/terms/
FaBiO: FRBR-aligned bibliographic ontology purl.org/spar/fabio fabio http://purl.org/spar/fabio/
FRBRcore purl.org/vocab/frbr/core frbr http://purl.org/vocab/frbr/core#
FRBRextended purl.org/vocab/frbr/extended# frbre http://purl.org/vocab/frbr/extended#
IFLA’s FRBRer Model iflastandards.info/ns/fr/frbr/frbrer/ frbrer http://iflastandards.info/ns/fr/frbr/frbrer/
International Standard Bibliographic
Description (ISBD)
iflastandards.info/ns/isbd/elements/ isbd http://iflastandards.info/ns/isbd/elements/
Lexvo.org Ontology lexvo.org/ontology lvont http://lexvo.org/ontology#
MARC Code List for Relators id.loc.gov/vocabulary/relators mrel http://id.loc.gov/vocabulary/relators/
Open Provenance Model Vocabulary purl.org/net/opmv/ns opmv http://purl.org/net/opmv/ns#
PRISM: Publishing Requirements for Industry
Standard Metadata
prismstandard.org prism http://prismstandard.org/namespaces/basic/2.0/
Provenance Vocabulary Core Ontology purl.org/net/provenance/ns prv http://purl.org/net/provenance/ns#
RDA Relationships for Works, Expressions,
Manifestations, Items
rdvocab.info/RDARelationshipsWEMI rdarel http://rdvocab.info/RDARelationshipsWEMI
Schema.org schema.org schema http://schema.org/

Aggregators
International coverage and diverse scope
◦ European digital heritage gateway Europeana
◦ DRIVER
◦ OpenAIRE
Compatibility with aggregators
◦ Important for repositories
◦ Common requirement for repositories
◦ Metadata have to meet specific criteria and adopt specific vocabularies
LOD adoption is the prevailing approach
◦ Brings an order to the chaos of disparate solutions

Benefits by LOD Adoption
Avoid vendor lock-ins
Allow complex queries to be evaluated on the results
◦ Utilize the full capacities of SPARQL
Content can be harvested and integrated by third-parties
◦ Ability to create meta-search repositories
◦ Researchers can browse, search and retrieve content from these repositories
Bring existing content into the Semantic Web
◦ New capabilities are opened

Synchronous Vs. Asynchronous Exports
SPARQL-to-SQL translation in the digital repositories
◦ Asynchronous approach seems more viable
◦ Real-time results may not be as critical
◦ RDF updates could take place in a manner similar to search indexes
◦ The trade-off in data freshness is largely remedied by the improvement in the
query answering mechanism
◦ Data freshness can be sacrificed in order to obtain much faster results
◦ Exposing data periodically comes at a low cost
◦ Information does not change as frequently as e.g., in sensor data
◦ Data is not updated to a significant amount daily
◦ Selection queries over the contents are more frequent than the updates

From DSpace to Europeana (1)
DSpace cultural heritage repository
Data model
◦ Dublin Core
◦ Europeana Data Model (EDM)
The problem
◦ How to transform item records as RDF using the EDM model

Components
◦ Source
◦ The relational database
◦ Target
◦ An RDF graph
◦ The R2RML Parser
Information flow
◦ Parse database contents into result sets
◦ Generate a Java object
◦ Instantiates the resulting RDF graph in-memory
◦ Persist the RDF graph
Parser GeneratorSource
database
RDF graphR2RML Parser
Mapping
file Hard disk

Bibliographic record example
Metadata field Metadata value
dc.creator G.C. Zalidis
A. Mantzavelas
E. Fitoka
dc.title Wetland habitat mapping
dc.publisher Greek Biotope-Wetland Centre
dc.date 1995
dc.coverage.spatial Thermi
dc.type Article
dc.rights http://creativecommons.org/licenses/by/4.0/

Output description (RDF/XML abbreviated)
<edm:ProvidedCHO rdf:about="http://www.example.org/handle/11340/615">
<dc:creator rdf:resource="http://www.example.org/persons#G.C. Zalidis"/>
<dc:creator rdf:resource="http://www.example.org/persons#A. Mantzavelas"/>
<dc:creator rdf:resource="http://www.example.org/persons#E. Fitoka"/>
<dc:title>
Wetland habitat mapping
</dc:title>
<dc:publisher rdf:resource="http://www.example.org/publishers#Greek Biotope-
Wetland Centre"/>
<dc:date>1995</dc:date>
<dcterms:spatial rdf:resource="http://www.example.org/spatial_terms#Thermi"/>
<dc:type rdf:resource="http://www.example.org/types#Article"/>
<dc:rights>
http://creativecommons.org/licenses/by/4.0/
</dc:rights>
</edm:ProvidedCHO>

DSpace relational database schema
◦ Basic infrastructure
◦ Allows arbitrary schemas and vocabularies
metadataschemaregistry
PK metadata_schema_id
namespace
short_id
metadatavalue
PK metadata_value_id
FK1 item_id
FK2 metadata_field_id
text_value
text_lang
place
authority
confidence
item
PK item_id
in_archive
withdrawn
last_modified
owning_collection
metadatafieldregistry
PK metadata_field_id
FK1 metadata_schema_id
element
qualifier
scope_note
handle
PK handle_id
handle
resource_type_id
resource_id

Triples Maps definitions in R2RML
Create URIs based on metadata values from Dspace
◦ Example: dc.coverage.spatial
◦ Subject (rr:subjectMap template)
◦ ' http://www.example.org/handle/{"handle"} '
◦ Predicate (rr:predicate value)
◦ dcterms:spatial
◦ Object (rr:objectMap template)
◦ ' http://www.example.org/spatial_terms#{"text_value"} '

map:dc-coverage-spatial
rr:logicalTable <#dc-coverage-spatial-view>;
rr:subjectMap [
rr:template
'http://www.example.org/handle/{"handle"}';
];
rr:predicate dcterms:spatial;
rr:objectMap [
rr:template
'http://www.example.org/spatial_terms#{"text_value"}';
rr:termType rr:IRI
];
].
<#dc-coverage-spatial-view>
rr:sqlQuery """
SELECT h.handle AS handle, mv.text_value AS
text_value
FROM handle AS h, item AS i, metadatavalue AS mv,
metadataschemaregistry AS msr, metadatafieldregistry
AS mfr WHERE
i.in_archive=TRUE AND h.resource_id=i.item_id AND
h.resource_type_id=2 AND
msr.metadata_schema_id=mfr.metadata_schema_id AND
mfr.metadata_field_id=mv.metadata_field_id AND
mv.text_value is not null AND i.item_id=mv.item_id
AND
msr.namespace='http://dublincore.org/documents/dcmi-
terms/'
AND mfr.element='coverage' AND
mfr.qualifier='spatial'
""".
R2RML mapping

Technical vs. Bibliographic dimension
Widespread ontologies have to be used where applicable
Linking the data to third party datasets using other datasets’
identifiers is also an aspect

Outline
Introduction
Motivation-Benefits
Complete example
Future outlook

Challenges: Ontology-based Data Updates
SPARQL-based access to the contents of the database is
unidirectional
Transform SPARQL Update requests to appropriate SQL
statements and execute them on the underlying relational
database
An issue similar to the classic database view update
problem

Challenges: Mapping Updates
Database schemas and ontologies constantly evolve
◦ Established mappings should also evolve, not be redefined or rediscovered
from scratch
An issue closely related to the previous one
Modifications in either participating model do not incur adaptations
to the mapping but cause some necessary changes to the other model
Could prove useful in practice
◦ Database trigger functions
◦ The Link Maintenance Protocol (WOD-LMP) from the Silk framework

Challenges: Linking Data
Reusing popular Semantic Web is not sufficient for the
generation of 5-star Linked Data
◦ Database values should not only be translated to RDF literals
◦ Real-world entities that database values represent should be
identified and links between them should be established
Related tools
◦ RDF extension for Google Refine
◦ T2LD

Chapter 5
Generating Linked Data in Real-
time from Sensor Data Streams

Outline
Introduction
Fusion
The Data layer
Rule-based Reasoning
Complete Example

Problem Framework
Rapid evolution in ubiquitous technologies
Pervasive computing is part of everyday experience
◦ User input
◦ Information sensed by the environment
Parallel decrease of the price of sensors
IoT and M2M

Streamed Data (1)
Need for real-time, large-scale stream processing
application deployments
Data Stream Management Systems
◦ Managing dynamic knowledge
◦ Emerged from the database community
◦ Similar concern among the Semantic Web community

Streamed Data (2)
Numerous challenges
◦ Large scale
◦ Geographic dispersion
◦ Data volume
◦ Multiple distributed heterogeneous components
◦ Sensor, sensor processing and signal processing (including a/v) components
◦ Vendor diversity
◦ Need for automation

Data Stream Management Systems
Fill the gap left by traditional DBMS’s
◦ DBMS’s are not geared towards dealing with continuous, real-time
sequences of data
Novel rationale
◦ Not based on persistent storage of all available data and user-
invoked queries
Another approach
◦ On-the-fly stream manipulation
◦ Permanent monitoring queries

Context-awareness, IoT and Linked Data (1)
Context
Context-aware systems
◦ Knowledge of the environment in which they are acting
◦ Extract and use information from the environment in order to
adjust their functionality according to the incoming information
◦ Behavior according to the existing conditions
◦ Tightly connected to the IoT vision
Any information that can be used to characterize the situation of an entity. An entity is a
person, place, or object that is considered relevant to the interaction between a user and an
application, including the user and applications themselves

The IoT vision
◦ Aims at connecting (large numbers of) sensors deployed around
the world
◦ Focus shifts to automated configuration of filtering, fusion and reasoning
mechanisms that can be applied to the collected sensor data streams
Ubiquitous, pervasive, context-aware
◦ The ability of a system to “read” its environment and take
advantage of this information in its behaviour

Challenge
◦ Heterogeneity in systems
◦ Capture, store, process, and represent information
◦ Information is generated in large volumes and at a high velocity
Common representation format
◦ Making systems interoperable
◦ Allows information to be shared, communicated and processed

Linked Data
◦ An efficient approach towards filling in this gap
◦ A common, reliable and flexible framework for information
management
◦ Information homogeneity
◦ Facilitates information integration

Outline
Introduction: Problem Framework
Fusion
The Data layer
Complete Example

Information Fusion (1)
Fusion
◦ Leverages information meaning
◦ Partial loss of initial data may occur
◦ Several fusion levels
The study of techniques that combine and merge information
and data residing at disparate sources, in order to achieve
improved accuracies and more specific inferences than could be
achieved by the use of a single data source alone

Algorithm
◦ Online (distributed)
◦ Each node can take decisions based only on its perception of the world
◦ Each algorithm execution is based on the knowledge of only a local node or
a cluster of nodes
◦ Offline (centralized)
◦ There is a need of a central entity maintaining system-wide information
Fusion nodes can
◦ Act as a server (push)
◦ Harvest information (pull)

Information fusion vs integration
◦ Fusion takes place in the processing steps
◦ Integration refers to the final step
◦ The end user’s gateway to (integrated) access to the information

Fusion Levels
Signal level
◦ Signals are received simultaneously by a number of sensors
◦ Fusion of these signals may lead to a signal with a better signal-to-noise ratio
Feature level
◦ A perceptual component must first extract the desired low-level features from each
modality
◦ Typically represents them in a multidimensional vector
Decision level
◦ Combines information from multiple algorithms in order to yield a final fused decision
◦ May be defined by specific decision rules
EarlyfusionLatefusion

JDL Fusion Levels (1)
A process model for data fusion and a data fusion lexicon
Intended to be very general and useful across multiple
application areas
Identifies the processes, functions, categories of techniques,
and specific techniques applicable to data fusion

Process conceptualization
◦ Sensor inputs
◦ Source preprocessing
◦ Database management
◦ Human-computer interaction
◦ Four key subprocesses (following next)

Level 1 – Object Refinement
◦ Combines sensor data together to obtain a reliable estimation of
an entity position, velocity, attributes, and identity
Level 2 – Situation Refinement
◦ Attempts to develop a description of current relationships among
entities and events in the context of their environment

Level 3 – Threat Refinement
◦ Projects the current situation into the future to draw inferences
◦ E.g. about friendly and enemy vulnerabilities, threats, and opportunities for
operations
Level 4 – Process Refinement
◦ A meta-process which monitors the overall data fusion process
◦ Assesses and improves the real-time system performance

Level 5 – Cognitive or User Refinement
◦ Added in revisions
◦ Introduce the human user in the fusion loop
◦ The aim is to generate fusion information according to the needs of the system user
National
Distributed
Local
INTEL
EW
SONAR
RADAR
.
.
.
Databases
Level 0
Subobject data
refinement
Level 1
Object
refinement
Level 2
Situation
refinement
Level 3
Impact
assessment
Level 4
Process
refinement
Human-
computer
interaction
Users
Database management system
Support
database
Fusion
database
Data fusion domain

Outline
Fusion
The Data layer
Complete Example

The Data Layer (1)
Metadata of multimedia streams
Semi-structured vs structured
◦ No significant technological challenges into converting semi-structured
documents into RDF
Data is produced in the form of streams
◦ Originating from a set of heterogeneous distributed sources
◦ No starting or ending point
◦ A strategy must be devised
◦ Define how new facts will be pushed into the system and old facts will be pushed out of it

The Data Layer (2)
Data flow
◦ Use of a middleware
◦ Semantic annotation → Knowledge Base
◦ Ability to answer semantic queries

Modeling Context (1)
Challenges
◦ Complexity of capturing, representing and processing the concepts
◦ Expressivity of the description of the world
◦ Expressive enough in order to enable specific behaviors
◦ Not as complex as to render the collected information unmanageable
Common representation format and vocabulary
◦ Ensure syntactic and semantic interoperability in the system
◦ Enable integration with third party data sources

Modeling Context (2)
Uniform context representation and processing at the
infrastructure level
◦ Better reuse of derived context by multiple data producers and
consumers
Ontology-based descriptions
◦ Offer unambiguous definitions of the concepts and their relationships
◦ Allow further exploitation of the created Knowledge Base
◦ Higher level, intelligent, semantic queries
◦ Formalize knowledge about the specific domain

Semantic Web Technologies in Sensor Networks
Research focuses on modelling
Goal: enable higher level processing for event/situation analysis
◦ Define ontologies for sensor measurement and sensor description
◦ Use a meta-ontology (e.g. SUMO)
◦ Define a sensor ontology under it
◦ Use it to annotate, query, and discover sensors
◦ Base concepts: Resource, Actor, Environment
◦ Parts that describe the state of the resources
◦ Entities operating in the resources
◦ Surrounding environment conditions

Semantics for Situation Awareness (1)
Situation Theory Ontology (STO)
◦ A unified expression of the Situation Theory
◦ Models events/objects and their relationships in OWL
◦ Can be extended with classes and relations that correspond to the
actual application scenario
◦ Two additional ontologies integrated, in order to be able to use it
in a real sensor fusion environment
◦ The Time ontology
◦ The WGS84 Geo Positioning ontology

Semantics for Situation Awareness (2)
SSN ontology
◦ Describes sensor data in the Linked Sensor Data context
Semantic Sensor Web (SSW)
◦ Framework for providing meaning for sensor observations
◦ Bridges the gap between
◦ XML-based metadata standards of the Sensor Web Enablement
◦ RDF/OWL-based vocabularies driving the Semantic Web

Modeling Context using Ontologies
A Local-As-View setting
◦ Optimal when new types of sensors need to be integrated and there is a
relatively high variety in structure
◦ Preferred approach
◦ Better than translating user queries from the global schema to each one of the local ones in
order to retrieve, aggregate and return results to the user
Data produced by sensors is eventually stored in its Knowledge Base
◦ Reasoning
◦ Infer knowledge which corresponds to events
◦ Situation assessment
◦ Cannot be performed at lower fusion levels

Sensor Data
Audiovisual
◦ Audio/video stream
Non-audiovisual
◦ Any kind of streamed sensor observations
◦ Temperature, RFID tags, etc.
Incoming data is subject to filtering
◦ Not all information is stored and processed
◦ Only the subset that is meaningful and needed by the processing modules
◦ E.g. discard out-of-range values

Challenges in Homogenizing Sensor Data
Numerous standards to annotate various areas of sensor data
◦ MPEG-7 for audiovisual
◦ Federal Geographic Data Committee (FGDC) for geographical information
Accuracy of the annotations
Convenience of updates and maintenance
Added value to the content itself
◦ Not always clear how the user can benefit from the existence of such annotations
Tracker errors
◦ False, missing, incorrect annotations

A Two-step Approach
Annotate sensor data according to the nature of its tracker
Homogenize the data under a common vocabulary
◦ Containing the captured values and semantics

Real-time vs. Near-real-time (1)
A real-time system
◦ Must satisfy explicit bounded response time constraints to avoid failure and
present consistency regarding the results and the process time needed to
produce them
◦ Emphasis in predicting the response time and the effort in reducing it
◦ Response time: the time between the presentation of a set of inputs and the appearance of
all the associated outputs
A real-time sensor fusion system
◦ Produces certain alerts (outputs)
◦ Connected to the appearance of certain inputs (events)

Real-time vs. Near-real-time (2)
Near real-time
◦ E.g. systems that schedule their operations at fixed time intervals
◦ Frequency of the updates depends on the application
◦ E.g. in a surveillance scenario, more frequent updates would be needed than in an
environmental monitoring scenario
Sensor inputs/outputs
◦ Real-time signals
◦ E.g. audiovisual streams
◦ Near-real time/asynchronous messages
◦ E.g. RFID readings

Data Synchronization and Timestamping (1)
Synchronization
◦ Rules of the form
if event1 occurred before event2 then…

Two kinds of timestamps
◦ The time the event was recognized
◦ Local time at the node that made the measurement
◦ Problem: how to synchronize the sensor network to a common clock
◦ The time the event arrived in the fusion node
◦ Fusion node
◦ A node that fuses information incoming from other nodes
◦ No need of a common clock
◦ The fusion node will timestamp events upon their arrival
◦ Followed when there are no great delays in communicating messages

Timestamping can be
◦ Distributed
◦ At each node
◦ Centralized
◦ At a central node, maintaining a common clock

Windowing
A mechanism to assure continuous data processing
In order to process newly generated information properly,
the system will not have to take into account all existing
information
Maintain a working memory window
Streams are unbounded
◦ Cannot fit into memory in order to be processed as a whole

Windowing-related Decisions
The measurement unit
The size
The window behavior
◦ Sliding windows
◦ Tumbling windows
◦ Landmark windows
◦ Partitioned windows
◦ Predicate windows
Rules applied real-time are restricted to the current information window

The (Distributed) Data Storage Layer (1)
Generated information needs to be stored and processed
before being communicated to the system
Each node maintains its perception of the real world
◦ Physically stored in a local database
Multi-sensor stream processing systems purposed to
function under a heavy load of information
◦ Preferred database design geared towards scalability
◦ I.e. decentralized to the maximum extent possible
◦ Not centralizing collected information

The (Distributed) Data Storage Layer (2)
An approach in order to maintain scalability
◦ Each node keeps the amount of information required for its local
operation
◦ Local database schema has to relate only to the hosted components
◦ Restrict information communicated throughout the system
◦ Communicates to the central node only higher level information
◦ E.g. detected events or entities

Relational to RDF in Sensor Data Streams
Produced messages will have to be eventually converted to
RDF and ultimately inserted in an ontology
A mapping layer
◦ Map the relational schema to the semantic schema

Mapping Layer (1)
Push strategy
◦ Forward data to the ontology using semantic notation as soon as
they are generated
◦ Advantages
◦ Transformations are executed fast
◦ The ontology is always up-to-date
◦ Disadvantages
◦ Each lower level node will have to implement its own push method
◦ Risk that the ontology will be populated with data even when no query is
sent to the semantic layer

Mapping Layer (2)
Pull strategy
◦ Transform relational data to semantic on request
◦ I.e. during query time
◦ Similar to RDF Views
◦ Advantages
◦ Actual mapping defined at semantic level
◦ Data transformed on request
◦ The ontology will accumulate instances needed for the actual query evaluation
◦ Disadvantages
◦ Could lead to longer response times during queries

Outline
Fusion
The Data layer
Complete Example

Rule-based Stream Reasoning in Sensor Environments (1)
Reasoning
◦ Infer implicit information
◦ Use the ontology structure and instances in order to draw conclusions
about the ongoing situations
◦ Based on a set of provided rules, applied on the created knowledge base
Extend the knowledge base by using rules
◦ To describe complex situations/events
◦ To create alarm-type of objects if certain conditions are met
◦ To depict the desired behavior and intelligence

Event-condition-action pattern
An event is a message arrival indicating a new available
measurement
Two distinct sets of rules
◦ Mapping rules
◦ Semantic rules
on event if condition then action

Mapping rules
◦ Specify how sensor measurements represented in an XML-based
format will be mapped to a selected ontology
◦ Can fetch data from the XML-like message and store it into the
ontology model in the form of class instances
◦ Can be perceived as the necessary step bridging the gap between
semi-structured data and ontological models

Semantic rules
◦ Derive new facts, actions or alerts
◦ Based on existing facts and knowledge
◦ Perform modifications on the ontology model
◦ Depend on the specific domain or deployment scenario
◦ Involve high-level concepts , meaningful to humans
◦ E.g., “when a certain area under observation is too hot, open the ventilating
system”
◦ The “too hot” conclusion will probably be inferred from a set of current
observations coupled with the knowledge stored in the ontology

Rule-based systems
◦ Combine real-time data with stored sensor data
◦ Fire rules based on data obtained from sensors in real-time and
classified as ontology instances
Rule-based problem solving
◦ An active topic in AI and expert systems
◦ Classic approach comes from work on logical programming and
deductive databases
◦ Conventional rule engine implementations based on the Rete algorithm

Stream reasoning
Will lead the way for smarter and more complex
applications
◦ E.g. traffic management, fastest route planning, environmental
monitoring, surveillance, object tracking, disease outburst
detection, etc.
Performing reasoning on a knowledge base comprising stable or occasionally
changing terminological axioms and a stream of incoming assertions or facts

RIF – Rule Interchange Format
◦ W3C recommendation
◦ A core rule language
◦ A set of extensions (dialects) that allow the serialization and
interchange of different rule formats
Also RuleML and SWRL
Also using SPARQL CONSTRUCT
◦ SPIN being an extension to this approach

Rule-based Reasoning in Jena (1)
Jena Semantic Web Framework
◦ The most popular Java framework for ontology manipulation
◦ Includes an inference engine, can be used as a reasoner
◦ Includes a number of predefined reasoners
◦ Transitive reasoner
◦ RDFS rule reasoner
◦ OWL, OWL mini, OWL micro
◦ DAML micro reasoner
◦ A generic rule reasoner that can be customized to meet specific ad hoc application
demands

Forward chain rules (body → head)
◦ When the body is true, then the head is also true
Built-in rule files of the form:
Symmetric and transitive properties in OWL:
[symmetricProperty1: (?P rdf:type owl:SymmetricProperty), (?X ?P ?Y) -> (?Y ?P ?X)]
[transitivePropery1: (?P rdf:type owl:TransitiveProperty), (?A ?P ?B), (?B ?P ?C) -> (?A ?P ?C)]
[rdfs5a: (?a rdfs:subPropertyOf ?b), (?b rdfs:subPropertyOf ?c) -> (?a rdfs:subPropertyOf ?c)]

Builtin primitives
◦ Functions that can be used in the place of rule predicates
◦ isLiteral(?x), notLiteral(?x),
◦ isFunctor(?x), notFunctor(?x)
◦ isBNode(?x), notBNode(?x), etc.
◦ Custom builtin primitives can be developed

Rule-based Reasoning in Virtuoso (1)
Reasoning essentially a set of rules applied on the RDF graph
Relatively simple reasoning
◦ Scalability instead of rich inference capabilities
Rule sets
◦ Loaded using the rdfs_rule_set function
◦ User can specify a logical name for the rule set, plus a graph URI
◦ Are provided as a context to user queries
◦ Can be referenced by SPARQL queries or endpoints

Rule-based Reasoning in Virtuoso (2)
Queries return results as if the inferred triples were
included in the graph
◦ Inferred triples generated by reasoning (the rule set) are generated
at runtime, are not physically stored

Outline
Fusion
The Data layer
Complete Example

Complete Example
Information originating from a distributed sensor network
Architecture of a Multi-Sensor Fusion System
◦ Based on GSN
◦ Operating at all JDL levels
◦ Create Linked Data
Fusion and its potential capabilities
Combine semantic web technologies with a sensor network middleware
Blend ontologies with low-level information databases

Proof-of-Concept Implementation
An LLF node
Two processing components
◦ A Smoke Detector
◦ A Body Tracker
◦ Each component hosted on a computer with a camera
◦ RTP streams
An HLF node
A Central node
◦ Overall system supervision

The GSN Middleware (1)
Needed in order to perform LLF
Open-source, java-based
Allows processing data from a large number of sensors
Covers LLF functionality requirements in sensor data
streams

Virtual sensor
◦ Any data provider (not only sensors)
◦ Configuration file in XML
◦ Processing class
◦ Windowing
◦ Time- or tuple-based sliding window size
◦ Data source
◦ Output fields

GSN Servers
◦ Can communicate between them
◦ Support input from more than one data stream
◦ Can form a network
◦ Allow information collection, communication, fusion, integration
Data acquisition
◦ An SQL-like procedural language
◦ Combine and fuse the information
◦ User-defined LLF functions
◦ GSN takes care behind the scenes about crucial issues
◦ E.g. thread safety, synchronization, etc.

Low Level Fusion (1)
Camera generates an RTP feed with its perception of
the world
Feed processed by signal processing components
◦ The Body tracker
◦ Reports NumberOfPersons
◦ The Smoke detector
◦ Reports NumberOfSmokeParticles
◦ Both components
◦ Receive a POLL command at fixed time intervals
◦ Return an XML file

Sampling the video source takes place asynchronously
◦ Camera streams at 25 fps
◦ 500 ms between two consecutive polls suffices for LLF
Tracking a person is more demanding than detecting it
◦ Implies comparing consecutive frames
Asynchronous processing
◦ The camera fps rate not aligned with the produced message rate

Virtual sensor XML configuration
files
◦ Body tracker
◦ PK: primary key (auto-incremented integer)
◦ Timed (timestamp)
◦ NumberOfPersons (integer)
◦ Similarly for the smoke detector
◦ ExistenceOfSmoke (boolean)
BodyTracker
PK PK
timed
NumberOfPersons
<virtual-sensor name="BodyTracker">
…
<output-structure>
…
<field name="NumberOfPersons"
type="int" />
</output-structure>
…
<storage history-size="1m" />
</virtual-sensor>

LLF virtual sensor definition
◦ Two data providers (source streams)
◦ Produce events only when the fusion conditions are satisfied
◦ A person and smoke are detected, concurrently
SELECT source1. NumberOfPersons AS v1,
source2. ExistenceOfSmoke AS v2
FROM source1, source2
WHERE v1 > 0 AND v2 = "true"

Fusion
◦ Results provided by taking into account the inputs from both the
sensors
◦ More processing components, sensors, sensor types and more
complex fusion conditions can be integrated into the system
Low Level Fusion
◦ Allow only important information to be forwarded to the upper
layers
No semantic enrichment so far

A Sensor Fusion Architecture (1)
From theory to practice
Address matters that correspond to all JDL levels
Level 0
Subobject data
refinement
Level 1
Object
refinement
Level 2
Situation
refinement
Level 3
Impact
assessment
Level 4
Process
refinement
Users
Database management system
Node
databases
Central
database
Data fusion domain
Signal processing
components
Sensor
Layer
Tracking &
Tracing
High Level
Fusion Node Central Node
(Common
Operational
Picture)
Ontologies/Semantic
Fusion
Control Node
Low-Level
Fusion Node

Collected data is leveraged into meaningful and
semantically enriched information
◦ I.e. transformed from signals to higher level information
All levels of multi-sensor data fusion are applied to the data

JDL Level 0 – Subobject data refinement
◦ Signal Processing components operate at this level
◦ System collects observation data
◦ Sensory data input mostly from Electro-Optical sensors (e.g. cameras,
microphones, etc.)
◦ Virtualizes the inputs of the sensors
◦ Performs spatial and temporal alignment

JDL Level 1 – Object refinement
◦ System analyzes the collected data
◦ Extracts features, objects, and low level events
◦ Detected objects may include persons or vehicles
◦ Low level events may include movements
◦ Implementation can be based on e.g. GSN

JDL Level 2 – Situation refinement
◦ System aggregates and fuses the objects, events and the general
context in order to refine the common operational environment
◦ System is able to perform object fusion and tracking, and identify
situations in the system, in a manner that is not feasible by a
sensor alone

JDL Level 3 – Impact assessment
◦ Events and objects refined at JDL Level 2 are subsequently
analyzed, using semantically enriched information, resulting at the
inferred system state
◦ High Level Fusion (HLF, e.g., using Virtuoso’s rule engine) operates
at this level
◦ Fuses the information, enabling reasoning that can infer and assess potential
impacts
◦ E.g. situations and respective alerts

JDL Level 4 – Process refinement
◦ System can refine its operation by providing feedback to the
sensor layer
◦ Ontologies/Semantic fusion component
◦ Perform analysis on the system-wide high-level collected information in
order to infer events and potential risks and threats
◦ Hosted at the Central Node
◦ Monitor and curate the whole system

Level 5 – Cognitive or User Refinement
◦ A higher level, introduced in revised versions of the JDL model
◦ Control Node component
◦ Hosted at the Central Node
◦ Responsible for issuing commands back to the sensors
◦ Common Operational Picture component
◦ System front-end
◦ Provide a visualization of the system state
◦ Deployed sensors, detected objects, sensed events and inferred threats
◦ Allows for human-computer interaction

Two types of databases to support system operation
◦ Support databases
◦ Materialized as a node database, kept locally at each node
◦ Collected data are kept close to their source of origin, allowing for each node
to be configured and maintained according to its environment and system’s
needs
◦ Allow system to scale
◦ Central Node database
◦ Plays the role of the fusion database, where higher level fused information is
kept

High Level Fusion Example (1)
Introduction
◦ Ontologies
◦ Powerful means to describing concepts and their relations
◦ Entities, events, situations, etc.
◦ JDL levels 2 and 3
◦ Main information gathering point in the proposed architecture
◦ Reasoning
◦ Infer new knowledge
◦ Cannot be performed at the LLF level

The scenario
◦ Demonstrate inference capabilities of the proposed architecture
◦ Full use of the data chain
◦ From low level sensor data to high level complex event detection and
situation awareness/threat assessment
◦ Detect threats to public safety and associate an action plan
◦ I.e. detect persons and smoke in an area of interest, e.g. a petrol
station

Configure GSN to use Virtuoso as its backend at the HLF node
Develop RDF views in Virtuoso over the sensor data
Use the Situation Theory Ontology (STO)
◦ The STO:FocalSituation class marks significant situations that prompt
action by security personnel

Perform reasoning in Virtuoso
◦ Triple store populated with data from the Smoke detector and Body
tracker components
◦ Reasoning associates the smoke detection event with a criticality factor
◦ According to the events and geospatial information modeled in STO
The scheduler component in Virtuoso
◦ Update the triple store via RDF views from the sensor inference database
◦ Subsequently invoke on the reasoning

Query to retrieve focal situations with their event related details
◦ Time, location, textual descriptions
SELECT ?focalsituation ?timeTxt ?locTxt
WHERE {
?focalsituation STO:focalRelation ?event .
?event STO:hasAttribute ?time .
?event STO:hasAttribute ?location .
?time rdf:type STO:Time .
?time time:inXSDDateTime ?timeTxt .
?location rdf:type STO:Location .
?location rdfs:label ?locTxt .
}

Integration with emergency departments achieved by
sending details on significant threats found in SPARQL result
sets to a Web service endpoint
◦ The smoke event together with location data is forwarded to
emergency personnel
More details in:
◦ Doulaverakis et al. An approach to intelligent information fusion in
sensor saturated urban environments. EISIC 2011, Athens, Greece

Chapter 6
Conclusions
Summary and Outlook

Outline
Introduction
Recap
Discussion
Open Research Challenges

Introduction
An envisioned Linked Data Interoperability Layer for
tomorrow’s Web
Physical Layer
Data Link Layer
Network Layer
Transport Layer
Session Layer
Presentation Layer
Application Layer7
6
5
4
3
2
1 Physical Layer
Data Link Layer
Network Layer
Transport Layer
Session Layer
Presentation Layer
Linked Data Layer
Application Layer
7a
7b
6
5
4
3
2
1

Outline
Introduction
Recap
Discussion

Chapter 1 – Introduction
Definitions – prospects – solutions
◦ Semantic Web
◦ Main building blocks
◦ Key terms
◦ Issues
◦ Linked Data

Chapter 2 – Technical Background
From theory to practice
◦ Introduction of the technical background that materializes Chapter
1 concepts
◦ Fundamental technologies
◦ From knowledge representation models to query languages and mappings
◦ Popular Linked Data vocabularies

Chapter 3 – Deploying Linked Data
A technical overview
Modeling data
Opening Data
Linking Data
Processing Data
◦ Available technical solutions and tools

Chapter 4 – Creating Linked Data from Relational Databases
RDBMS with Semantic Web applications interfaces
Motivations
Benefits
Related literature survey
◦ Approach categorization
Proof-of-concept use case
◦ Convert data from an open access repository to Linked Data

Chapter 5 – Generating Linked Data in Real-time from Sensor Data Streams
Basic concepts
◦ Introduction: real-time processing, context-awareness, windowing
and information fusion
Related Issues
System description
◦ An intelligent, semantically-enabled data layer to integrate sensor
information

Overall Contribution (1)
◦ Metadata
◦ Real-time
◦ Context-awareness
◦ Integration
◦ Interoperability
◦ Fusion, etc.
Formal and informal introduction of Data Science concepts
◦ Semantics
◦ Ontologies
◦ Data and Information
◦ Knowledge Bases
◦ Reasoning
◦ Annotation

Overall Contribution (2)
A detailed state-of-the-art survey
◦ Technologies, methodologies, tools, and approaches
Discussions on Linked Data creation
◦ From relational databases
◦ From sensor data streams
Two detailed architectural and behavioral
descriptions of two domain-specific scenarios

Outline
Introduction
Recap
Discussion

Discussion (1)
LOD ecosystem
◦ An open and distributed system
◦ Heterogeneity is inevitable
◦ At a syntactic, terminological, conceptual, or semiotic/pragmatic level
◦ The Linked Data paradigm
◦ Offers solutions to reduce heterogeneity at all four levels
◦ Defines relations across the heterogeneous sources

Discussion (2)
Server-side: many steps, many components involved
◦ No "standard" approach
◦ No deterministic manner in setting behavioral priorities
◦ The importance of scalability
◦ Volume of the produced, stored and processed information
◦ The number of the data sources
◦ The nature of the data
◦ E.g. sensor/multimedia/social network data streams

Discussion (3)
Integration with third parties
◦ Semantic Web technology adoption is crucial in order to assure
unambiguous definition of the information and the semantics it
conveys
◦ E.g. as in the example in the scholarly/cultural heritage domain
Data repository turned into a Knowledge Base
◦ Virtually endless possibilities
◦ E.g. analytical reasoning, intelligent analysis
◦ Data mining, pattern extraction, etc.
◦ Even in unprecedented ways

Benefits (1)
Increased discoverability
◦ Description of the dataset contents
◦ Links towards instances in other parts of the LOD cloud
◦ Inbound links are welcome
Reduced effort for schema modifications
◦ New relations allowed without modifying the database schema or
contents
◦ E.g. define new classes and properties

Benefits (2)
Synthesis
◦ Integration, fusion, mashups
◦ Allow searches spanning various repositories, from a single
SPARQL endpoint
◦ Allow download of parts or the whole data
Inference
◦ Reasoning support
◦ Implicit facts can be inferred, based on the existing ones, then
added to the graph

Benefits (3)
Reusability
◦ Third parties can reuse the data in their systems
◦ Including the information in their datasets, or
◦ By reference to the published resources

Technical Difficulties (1)
Multidisciplinarity
◦ Annotation task not to be underestimated
◦ Contributions required from several scientific domains
◦ Domain-specific expertise
◦ Close collaboration of the implementation team
Technology barrier
◦ Tools are not as mature yet as to provide guidance or warnings
◦ E.g. in the linking or mapping procedure
◦ No design-time validity of the result

Error-prone result
◦ Even syntactically correct
◦ No automatic check of whether the concepts and properties
involved are used as intended
◦ Errors or bad practices can go unnoticed
Concept mismatch
◦ Extraction of stored values into RDF not always possible
◦ Identical mappings may not always be found

Exceptions to the general rule
◦ Automated changes will apply to the majority of the data
◦ The remaining portion will require manual intervention
◦ Post-publishing manual interventions will be required

Open Government Data
Emerging, across governments and organizations from all over the world
◦ E.g. US, UK
Foster transparency, collaborative governance, innovation
Enhance citizens’ quality of life through the development of novel
applications
Data value decreases if not released in open formats, allowing
combination and linking with other open data
Ongoing efforts to integrate open governmental data to the LOD cloud

Bibliographic Archives (1)
A huge wealth of human knowledge exists in digital libraries
and open access repositories
Structured metadata
◦ Cataloguing, indexing, searching
◦ Typically trapped inside monolithic systems that support Web-
unfriendly protocols for data access
Linked Data
◦ Allows direct reuse of the work of other librarians
◦ Transforms the item-centric cataloguing to entity-based descriptions

Bibliographic Archives (2)
Several efforts at national and regional level
◦ E.g. Library of Congress, British National Bibliography
Linking of bibliographic data with LOD datasets from other
domains
◦ Expected to give rise to novel applications that exploit library data
in combination with other (e.g. geographical) data

Internet of Things
Billions of sensing devices currently deployed worldwide
◦ Already form a giant network of connected "things“
◦ Their number expected to continuously grow
◦ Uniquely identified and accessed through standard Internet protocols
Applications in diverse domains
◦ E.g. environmental monitoring, energy management, healthcare and home and
city automation
Intelligence in IoT
◦ Use of ontologies and other Semantic Web technologies that support inference
◦ Integration of developed independently deployed IoT platforms

Outline
Introduction
Recap
Discussion

Data Science
LOD provision is the first step
◦ Not a goal in itself; a means to an end
Consuming (as opposed to producing)
◦ Extract intelligence, generate additional value
◦ E.g. visualization, analytics, text mining, named entity recognition, etc.
◦ Quality assessment
◦ LOD quality ranges from extensively curated datasets to crowd-sourced and extracted data
of relatively low quality
The study of the generalizable extraction of knowledge from data

Big (Linked) Data (1)
Data in several media channels
◦ E.g. social networks, blogs, multimedia sharing services
◦ Generated in growing rates
◦ Influences professional and personal decisions and actions of
individuals
Size of the dataset is a part of the problem itself
◦ Millions or billions of facts
◦ Difficult to process using conventional data processing applications
◦ Storage and querying problems

No formal definition of what exactly Big Data is (and what is not)
◦ Commonly characterized by different properties
◦ All V’s for some mysterious reason
◦ Volume
◦ Velocity
◦ Variety
◦ Value
◦ Veracity

Need for
◦ Timely, accurate, efficient analysis of Big Data volumes
◦ Managing, querying and consuming
◦ Handling the vast amounts of data to be generated in the near
future
Linked Data
◦ Part of the Big Data landscape
◦ Ideal testbed for researching key Big Data challenges

(Even) More Research Challenges
Privacy
Legal aspects
Integration and reconciliation from diverse data sources

Materializing the Web of Linked Data

More Related Content

What's hot (18)

Viewers also liked (20)

Similar to Materializing the Web of Linked Data (20)

More from Nikolaos Konstantinou (8)

Recently uploaded (20)

Materializing the Web of Linked Data