How to Create Controlled Vocabularies for Competitive Intelligence

The Intelligence Collaborative
http://IntelCollab.com #IntelCollab
Powered by
How to Create Controlled
Vocabularies for
Competitive Intelligence
A Complimentary Webinar from Aurora WDC
12:00 Noon Eastern /// Wednesday 18 March 2015
~ featuring ~
Justin Soles Michel BernaicheLisa Coady

Powered by
The Intelligence Collaborative is the online learning and networking
community powered by Aurora WDC, our clients, partners and other friends
and dedicated to exploring how to apply intelligence methods to solve real-
world business problems.
Apply for a free 30-day trial membership at http://IntelCollab.com or learn
more about Aurora at http://AuroraWDC.com – see you next time!
Justin Soles
Justin Soles is an experienced CI analyst (and sometimes CI programmer,
instructor and evangelist), as well as a seasoned knowledge taxonomist and
information architect. Justin’s CI-related work has included patent and IP analysis,
competitor profiling/tracking and dashboard development for senior
management, among other projects. During his Masters in Library & Information
Studies at McGill University, Justin participated in the consolidation of several
Astronomy vocabularies into a single thesaurus along with experts from Harvard
University, the Smithsonian, the University of Edinburgh and various prestigious
scientific journal publishers.
Lisa Coady
Lisa Coady has 10+ years as a strategic development officer in the non-profit
sector. Having recently completed a Masters in Library & Information
Studies, she is setting up an information management consultancy
specializing in developing controlled vocabularies, taxonomies and thesaurii
to drive and enhance CI efforts in both commercial and non-profit
organizations.

Powered by
α Use the Questions pane on your
GoToWebinar control panel and all questions
will be answered in the second half of the
hour.
α You are welcome to tweet any comments on
Twitter where we are monitoring the
hashtag #IntelCollab or eavesdrop via
http://tweetchat.com/room/IntelCollab
α Slides will be available after the webinar for
embedding and sharing via
http://slideshare.net/IntelCollab
α To view the recording and download the PPT
file, please register for a trial membership at
http://IntelCollab.com.
Questions, Commentary & Content

Powered by
Agenda
► What is a Controlled Vocabulary (CV)?
► What’s the connection between CVs and CI?
► How do I create a CV for CI?
► What are some best practices/lessons
learned?
► Summary, Q&A and Discussion

Powered by
What is a Controlled Vocabulary?
► Some formal definitions:
► OECD: “…an established list of standardized
terminology for use in indexing and retrieval of
information.”
► Getty Institute: “…an organized arrangement of words
and phrases used to index content and/or to retrieve
content through browsing or searching.”
► SCIP: “… a standardized – yet dynamic – set of terms
and phrases authorized for use in an indexing system to
describe a subject area or Information domain.” (Vernon Prior)

Powered by
What does a CV look like?
List: Ungrouped set of
terms
Pepsi Subsidiaries:
- Tropicana Products, Inc.
- Frito Lay, Inc.
- Quaker Oats Ltd.
- The Gatorade Company
- Stokely Van-Camp, Ltd.
- Walkers Snack Foods Ltd.
- Green Hemlock
International, LLC
Hierarchy: Grouped set of
terms
Vehicles
► Land Vehicles
► 1-wheel vehicles
► Unicycle
► Bicycle
► Motorcycle
► Car
► Sports car
► Sedan
► SUV
► School bus
► Quad bike
► > 4 wheel vehicles
► Intercity bus

Powered by
List: Ungrouped set of
terms
Pepsi Subsidiaries:
- Tropicana Products, Inc.
- Frito Lay, Inc.
- Quaker Oats Ltd.
- The Gatorade Company
- Stokely Van-Camp, Ltd.
- Walkers Snack Foods Ltd.
- Green Hemlock
International, LLC
Common types of lists used
in CVs:
►Glossary/Lexicon:
List of domain-related
terms, usually with
definitions & usage
►Name Authority File:
List of approved names
for a related set of
objects

Powered by
Common types of
hierarchies used in CVs:
► Taxonomy: Hierarchy
showing the relationships
between terms
► Thesaurus: Taxonomy
with added context
(e.g. synonyms, preferred
terms, non-preferred
terms, scope notes)
Hierarchy: Grouped set of
terms
Vehicles
► Land Vehicles
► Unicycle
► Bicycle
► Motorcycle
► Car
► Sports car
► Sedan
► SUV
► School bus
► Quad bike
► > 4 wheel vehicles
► Intercity bus

Powered by
What kinds of (CI-related) information can you
organize with a CV?
Use a CV based on: To classify by/into: Example:
Document Type Document type of artifact
(beyond file type, date, etc.)
Type: Job Advertisement
Type: Press Release
Organizational Entity Publishing organization,
company, department, etc.
Entity: Human Resources
Entity: Marketing
Functional Process Process/Function to which
an artifact relates
Process: Staffing
Process: Promotion
Subject-based Subject area to which an
artifact pertains
Subject: Data Science
Subject: Donation to
University
Facet-based Different hierarchies, each
of which describes a
different aspect or piece of
information
Facet: Job Title - Data
Scientist, Skills – Hadoop, R
Languages – Python, PHP
Facet: Donation type -
Educational, Amount - $1M

Powered by
Where can CVs be used?
Back End: index & organize
- (Metadata) Tagging
- Textual analysis
- Document indexing
- Search engine
enhancement (result set
expansion via synonyms)
- Semantic/Contextual
understanding
(Ontology)
Front End: orient & inform
- Portal/Site organization &
navigation (Information
Architecture)
- Search engine
enhancement (auto-
complete, faceted search)

Powered by
How Controlled Vocabularies help the
Competitive Intelligence Cycle

Powered by
How Controlled Vocabularies help the
Competitive Intelligence Cycle
1. Establish a common
terminology framework
for KITs/KIQs
2a. Identify search
keywords
2b. Organize &
store search
results consistently
3a. Improve
artifact findability
4. Enhance impact
of CI deliverable
5. Improve
findability of
deliverables
3b. Facilitate
comparisons &
correlations

Powered by
How to develop a Controlled Vocabulary
for CI?
Preliminary
Activities
Discover
BuildTest
Maintain

Powered by
How to develop a Controlled Vocabulary
for CI?
Preliminary
Activities
Discover
BuildTest
Maintain
Define scope,
resources, budget
& timeline
Identify
sources &
collect terms
Create CV
(Some assembly
required)
Test CV
(Some reassembly
required!)
Add, edit &
retire terms

Powered by
Role of the Analyst
► Today’s solutions are successful in accessing business and
competitive data, and making it available to users, however,
.....today’s solutions are not so successful in directly linking
information to action, and deriving its corresponding value.
Dr. Craig Fleisher, presenting at SCIP African Summit Workshop @ UNISA
https://www.scip.org/re_pdfs/1415891934_pdf_SCIPAfricanSummitAnalysisMethodsWorkshpSessionSlidesbyDrCraigFleisherOct14.pdf
-------------------------------------------------------------------------------------------
The effective analyst bridges the gap between information and intelligence. (CF).
Controlled Vocabulary one of the analyst’s most valuable tools. In assigning data to
categories based on its attributes, you can make meaningful comparisons and draw
meaningful inferences within and about your domain (LC).

Powered by
How will we shape our Corporate
Social Responsibility strategy for
maximum effectiveness?
Scenario:
Using controlled vocabulary to drive a Corporate
Social Responsibility Analysis for a fuel extraction company

Powered by
Task: CSR assessment
► Task: perform a CSR assessment competitors
► Data-gathering
► Data analysis
► Development of a CSR framework for the company
► Assumptions:
► Distributed team
► (>1 person gathering, tagging, analyzing)
► Ongoing strategic priority (environment monitoring)

Powered by
►What is CSR?
► Corporate Social Responsibility (CSR) refers to voluntary activities undertaken by a
company to operate in an economic, social and environmentally sustainable manner.
( http://goo.gl/bTskh4 Foreign Affairs, Trade and Development Canada)
► Philanthropic or in-kind donations to social sectors, institutions, groups, causes
► Business Case

Powered by
Our “categorize to analyze” process
► Building a CV for business analyses is a data-
driven process
► Fundamental questions
► Collect sample data
► Scan and group according to shared attributes that
you assign through formal categorization
“categorize to analyze”
► Insights

Powered by
1. Fundamental questions
Scope
Analysis
► Who, what, where, when?
► KIQs
► Priority investment areas?
(where competitors are making CSR investments)
► Optimal funding mechanisms?
► Strategically aligned?

Powered by
2. Collect your data
Corp. Donor Recipient Year Amount CSR ”Instance" CSR Event Description
Company A XXX 2011 515,001 Lorem ipsum Name Lorem Ipsum Descripsum
Company A XXX 2011 1,000,000 Lorem ipsum Name Lorem Ipsum Descripsum
Company B XXX 2011 112,000 Lorem ipsum Name Lorem Ipsum Descripsum
Company B XXX 2013 1,000,000 Remote Medicine Technology for the
Medical School of University 2
Company B gives a one-time donation of $1M towards the
purchase of state-of-the-art medical visualization systems for the
medical school in University2. The donation is part of Company B's
commitment to support the training of the next generation of
health professionals, in the community in which it operates
Company B XXX 2010 2,100,000 Lorem ipsum Name Lorem Ipsum Descripsum
Company C XXX 2012 774,136 Lorem ipsum Name Lorem Ipsum Descripsum
Company C XXX 2014 1,200,000 Lorem ipsum Name Lorem Ipsum Descripsum
Company C XXX 2011 2,000,000 Lorem ipsum Name Lorem Ipsum Descripsum
CSR DATA - TELLS US
WHAT?
Reflects our scope
It’s a faceted taxonomy;
We can group on attributes and
make discoveries, just not many.
The data is largely unstructured.
CSR DATA –
DERIVED FROM WHERE?
CSR records found in
sustainability reports; filings of
corporate charitable foundations;
donor and recipient websites
Sample of N=420 CSR data points from X fuel companies in North America over past 4 yrs.
And compile
DOESN’T YET
ANSWER MY KIQs
(Where are my
competitors are investing
in CSR)

Powered by
3. Do your CV work
Corp. Donor Year Amount CSR "Event" Funding Mechanism Investment Area Strat Alignment?
Company A 2011 515,001 Lorem ipsum Name
Company A 2011 1,000,000 Lorem ipsum Name
Company B 2011 112,000 Lorem ipsum Name
Company B 2013 1,000,000 Remote Medicine Technology for the
Company B 2010 2,100,000 Lorem ipsum Name
Company C 2012 774,136 Lorem ipsum Name
Company C 2014 1,200,000 Lorem ipsum Name
Company C 2011 2,000,000 Lorem ipsum Name
INVESTMENT AREA?
FUNDING MECH?
ALIGNMENT?
Analysis
Formalize KIQs as facets
Begin developing CVs for facets
CV Terms? CV Terms? CV Terms?
Corp. Donor Year Amount CSR "Event" Funding Mechanism Investment Area Strat Alignment?
Company A 2011 515,001 Lorem ipsum Name Research Partnership Education/Innovation Yes
Company A 2011 1,000,000 Lorem ipsum Name Donation Community Development/Social Justice No
Company A 2012 120,000 Lorem ipsum Name In-Kind Contribution Community Development/Social Justice Yes
Company A 2013 475,975 Lorem ipsum Name Donation Environment/Preservation No
Company B 2011 112,000 Lorem ipsum Name Employee Matching Health/Foundation Support No
Company B 2013 1,000,000 Remote Medicine Technology for the
Research Partnership Education/Innovation Yes
Company B 2010 2,100,000 Lorem ipsum Name In-Kind Contribution Education/Innovation Yes
Company C 2012 774,136 Lorem ipsum Name Donation Health/Special Projects No
Company C 2014 1,200,000 Lorem ipsum Name Donation Health/Scholarhips No
Company C 2011 800,000 Lorem ipsum Name Employee Matching No
Company C 2012 50,000 Lorem ipsum Name Research Partnership Environment/Awareness Yes
Company C 2013 135,000 Lorem ipsum Name Employee Matching Environment/Preservation Yes
Company C 2011 2,000,000 Lorem ipsum Name Donation Environment/Innovation Yes
Enhanced filtering, increased granularity,
combinatorial power!
A series of attribute lists including one
hierarchy

Powered by
Vocabulary challenge?
►Must always design with
end-user’s perceptions
and language in mind
►Finding the balance in
representing your
domain using structure,
vocab and clear
definitions
► Ensuring oversight and
change mgmt.
Definition reveals the challenges
A controlled vocabulary is a collection of
terms that are:
Accepted: The term adheres to
community practices and reflects
community understanding
Defined: Vocabulary is adequate to cover
the domain. The terms are precisely
characterized with definitions. Imperative
for user understanding and for analytic
rigour.
Managed: A body of experts will create
and maintain the controlled vocabulary,
including: periodic review, modification of
terms and quality assurance.

Powered by
Sometimes it seems “simple”
Make no assumptions!
• About language, perspective or understanding of scope
• Provide a dictionary where the scope and meaning of
each term can be found (link)
“Simple” attribute?
• The specific attribute you are describing with your CV is
shallow (flat list)
• Limited #terms, widely understood and used
Funding Mechanism
Philanthropic Donation
Matching Funds
In-Kind
Research Partnership

Powered by
Often, it’s more complex
Structural Considerations
consider the golden rules of IA (use hierarchy, limit the
number of options per level, bear granularity in mind)
Investment Area
Investment Area is a more complex facet
How should we represent it
Vocabulary Considerations
consult subject matter experts (domain language?)
do a “physical build” of your taxonomy (card sorting)
draft a preliminary taxo
Test with end users (sorting, web taxo, observation)
Expect that it will be a long and iterative process

Powered by
Build, test and tweak (facet, investment area)
Education
Scholarships
Education Foundation Support
Special Projects
Innovation
Health
Health Training Scholarships
Health Foundation Support
Special Projects
Innovation
Disaster Relief
National
International
Environment
Awareness
Preservation
Innovation
Community Development
Economic Development
Social Justice
Social Innovation
Arts, Culture and Leisure
Food and Agriculture
Safety
Scenario A:
Company E gives a one-time donation of $1.5 M to buy
state-of-the-art medical visualization systems for the
medical school in University2.
What top-level term? Clarify your rules
Scenario B:
Company D includes supplies fuel for more than 180
emergency response vehicles in the local community where
it is headquartered.
Where? New term?
Investment Area

Powered by
Don’t look for the “right” term, seek
consensus and shared understanding
► With that, you have the power of efficient
organization, enhanced retrieval and the ability
to derive insights from your data
► In our case study, this may include insights that
can be compiled in a CSR dashboard:

Powered by
Corporate CSR, by year and IA
0
2000000
4000000
6000000
8000000
10000000
12000000
14000000
2010 2011 2012 2013 2014 2015
CSR Allocation by Year and Investment Area
Community Development Culture, Arts and Leisure Education Environment Health Safety
I want to look at this only for
companies in my province. Filter
by location

Powered by
Who spends most?
0
2000000
4000000
6000000
8000000
10000000
12000000
Company A Company B Company C Company D Company E Company F Company G Company H
Total CSR Expenditure by Corporation
Total
I want to see the yearly trend.
Combine with year facet.

Powered by
Strategic Alignment
No
40%
Yes
60%
Is CSR Strategically Aligned?
I want to see if strategic
alignment varies over
investment area.
Combine.

Powered by
Thank you! Now how about a little Q&A?
Email: justin.soles@mail.mcgill.ca
The Intelligence Collaborative is the online learning and networking
community powered by Aurora WDC, our clients, partners and other friends
and dedicated to exploring how to apply intelligence methods to solve real-
world business problems.
Apply for a free 30-day trial membership at http://IntelCollab.com or learn
more about Aurora at http://AuroraWDC.com – see you next time!
Justin Soles
Lisa Coady
Email: lisa.coady@mail.mcgill.ca

Powered by
Still time to register…

How to Create Controlled Vocabularies for Competitive Intelligence

More Related Content

What's hot (20)

Viewers also liked (11)

Similar to How to Create Controlled Vocabularies for Competitive Intelligence (20)

More from IntelCollab.com (20)

Recently uploaded (20)

How to Create Controlled Vocabularies for Competitive Intelligence