Contentdriven Knowledge Management through Evolving Ontologies

1

• Content-driven Knowledge Management through
  Evolving Ontologies
• On-To-Knowledge IST-1999-10132
• KM Europe 2000, 20-22 November,
• Palais des Congres, Brussels, Belgium
• York Sure Robert Engels,
• Institute AIFB, University of Karlsruhe
• CognIT a.s, Norway

York Sure Robert Engels
2
Agenda
Project partners
On-To-Knowledge
and their tasks
3
IST Information societytechnologies

• Fundamental transformations
• from an industrial society to the information
  society
• from stand-alone systems to networked
  information and processes
• Increasing diversity and complexity of systems
  presents new challenges for their
  development and use

4
IST Information societytechnologies

• Programme objective
• A user-friendly
• information society
• Realise the benefits of the information society
  for Europe by ..
• .. accelerating its emergence
• .. ensuring that the needs of individuals and
  enterprises are met

5
On-To-KnowledgeFacts

• European project in EU Information Society
  Technologies Programme IST-1999-10132
• Aim innovative tools for knowledge management
• Duration 2.5 years, January 2000 - June 2002
• Total effort cost 26 personyears, 2.5 M EUR
• EC funding 1.34 M EUR

6
Motivation

• IT, and especially Internet/WWW, have boosted
  potential for knowledge acquisition and sharing

• BUT information resources are heterogeneous,
  distributed, semi-structured, enormous in size
• HENCE need for KM tools for selective semantic
  (meaning-oriented) access gt On-To-Knowledge

7
Goals

• The information client
• Improve access to information sources!
• The information provider
• Lower costs for providing and maintaining large
  bodies of textual and semistructured information!

8
Workpackages

• Tools for Ontology Construction and
  Interchange
• Infrastructure for Information representation
• Various means for Information access
• Support for Information Extraction

• Methodology and Management Guidelines

• A case study on Call centres
• A case study on Knowledge management in a
  virtual enterprise
• A case study on Organisational memory

• Information dissemination and take-up of results
• Project Management and Exploitation Management

9
On-To-Knowledge architecture
Case Study Organisational Memory, SwissLife
Case Study Call Center, British Telecom
Case Study Virtual Enterprise, EnerSearch

• Goal-oriented methodology for knowlege
  management, AIFB

Uses
Access Layer, BT
Ontologies, VU
Representation Layer, AIdministrator
Represents
Uses
Extraction Layer, CognIT
10
What are ontologies?

• Describe a formal, shared conceptualization of
  particular domains of interest
• Describe
• concepts relevant for the domain,
• their relationships, and
• axioms about these concepts and relationships
• Enforces a well-defined semantics on such a
  conceptualization

• Aid to human communication and shared
  understanding, by specifying meaning
• Machine-processable (e.g., agents use ontologies
  in communication)
• Ontology key technology in semantic information
  processing

11
Vrije Universiteit Amsterdam, NL

• Free University NL, about 14.000 students
• Research power and achievements
• Knowledge Management (e.g. CommonKADS
  Methodology, Akkermans v.Harmelen UvA)
• e-business (HomeBot Agents, e-markets, e3-valueÒ)
• International business and industry network
• Strong EU-IST record RD, transfer, and
  management

12
Vrije Universiteit Amsterdam, NL

• Tools for Ontology Construction and
  Interchange
• Infrastructure for Information representation
• Various means for Information access
• Support for Information Extraction

• OIL for the Semantic Web
• Definition of language
• Semantics
• XML encoding
• RDF encoding
• Ontology construction, evolution and mapping

• Methodology and Management Guidelines

• A case study on call centres
• A case study on knowledge management in a virtual
  enterprise
• A case study on Organisational memory

• Information dissemination and take-up of results
• Project Management and Exploitation Management

13
AIdministrator, NL

• Software products that exploit intelligent
  techniques for information management
• Reduce human effort needed for
• searching, selecting, combining, maintaining
  information sources
• Small team aimed at intensive technology
  development

14
AIdministrator, NL

• Tools for Ontology Construction and
  Interchange
• Infrastructure for Information representation
• Various means for Information access
• Support for Information Extraction

• Key idea
• website hyperlinks ontology meaningful
  structure

• Methodology and Management Guidelines

• A case study on call centres
• A case study on knowledge management in a virtual
  enterprise
• A case study on Organisational memory

• Information dissemination and take-up of results
• Project Management and Exploitation Management

15
CognIT a.s, NO

• Robs presentation!

16
Institute AIFB, DE

• University of Karlsruhe, Research group
  Knowledge Management
• Knowledge Engineering
• Knowledge Portals
• Intelligent Web Applications -- Semantic Web
• Text Mining
• Ontology Learning from Text
• Ontology-based Information Extraction
• Semi-Automatic Text Annotation
• Data Mining
• Analysis of very large data sets

17
Institute AIFB, DE

• Tools for Ontology Construction and
  Interchange
• Infrastructure for Information representation
• Various means for Information access
• Support for Information Extraction

• Methodology and Management Guidelines

• A case study on call centres
• A case study on knowledge management in a virtual
  enterprise
• A case study on Organisational memory

• Information dissemination and take-up of results
• Project Management and Exploitation Management

18
Institute AIFB Methodology

• Identification of knowledge needs, opportunities
  and bottlenecks in focus areas of the
  organisation
• Structured construction and validation of
  ontologies
• User-focused introduction of KM tools
• Maintenance and steady improvement of existing KM
  applications

Context-oriented requirements for envisaged
knowledge management tools Principled foundation
for any knowledge management tool in the
organisation Knowledge pull and push instead of
information overload KM applications remain
successful and useful in a changing
organisational environment
Context-oriented requirements for envisaged
knowledge management tools Principled foundation
for any knowledge management tool in the
organisation Knowledge pull and push instead of
information overload KM applications remain
successful and useful in a changing
organisational environment
19
British Telecom, UK

• Public limited company registered in England
• Listings on the London, New York and Tokyo
  stock exchanges
• Vision to be the most successful worldwide
  communications group
• BT Cellnets customer connections up by 2.9
  million to 7.4 million
• BT is developing a range of products designed for
  efficient and effective knowledge management
• Come and visit BTs booth at B40!

20
British Telecom, UK

• Tools for Ontology Construction and
  Interchange
• Infrastructure for Information representation
• Various means for Information access
• Support for Information Extraction

Hopefully you didnt miss it today John
Davies Knowledge Tools for the Intranet

• Methodology and Management Guidelines

• A case study on call centres
• A case study on knowledge management in a virtual
  enterprise
• A case study on Organisational memory

• Information dissemination and take-up of results
• Project Management and Exploitation Management

21
EnerSearch AB, SE

• RD Consortium for utility business strategies
• Present shareholders
• ABB Automation Products AB, ECN Netherlands
  Energy Research Foundation, EDP Electricidade
  Portugal, Iberdrola, IBM Utility Services, E.ON
  energie, Sydcraft AB

22
EnerSearch AB, SE

• Tools for Ontology Construction and
  Interchange
• Infrastructure for Information representation
• Various means for Information access
• Support for Information Extraction

• Case study
• How to do knowledge transfer via website?
• Issue hyperlinks or keyword search do not tell
  you much, and are a waste of time
• Much better you want to ask questions and just
  get the answer

• Methodology and Management Guidelines

• A case study on call centres
• A case study on knowledge management in a virtual
  enterprise
• A case study on Organisational memory

• Information dissemination and take-up of results
• Project Management and Exploitation Management

23
SwissLife, CH

• Financial services company with strong focus on
  life insurance, long-term savings and asset
  management
• Largest and longest established life insurance
  company in Switzerland
• Operations in major European countries and
  representative offices in China and Australia
• Market leader in the domestic life insurance
  market (1999 market share of around 30) and no.
  8 in Europe
• Swiss Life Network with leading insurance
  companies world-wide
• 11500 employees world-wide

24
SwissLife, CH

• Tools for Ontology Construction and
  Interchange
• Infrastructure for Information representation
• Various means for Information access
• Support for Information Extraction

• Case studies
• Providing Knowledge about International
  Accounting Standard (IAS)
• Skills Management

• Methodology and Management Guidelines

• A case study on call centres
• A case study on knowledge management in a virtual
  enterprise
• A case study on Organisational memory

• Information dissemination and take-up of results
• Project Management and Exploitation Management

25
SwissLife Vision of an OM
26
SwissLife, CH
Case Study Providing Knowledge about
International Accounting Standard (IAS)

• Trade of Swiss Lifes shares on the European
  stock markets
• Every subsidiary must provide figures as required
  by IAS
• IAS deviates from the countries local standards
• Not much knowledge yet available inside Swiss Life

27
SwissLife, CH
28
SwissLife, CH
Providing Knowledge about IAS / Approach

• natural-language sentences as search expressions
  (CognIT a.s)
• automatically map to IAS terminology (use
  ontology)
• integrate FAQs into searchable knowledge
• enable expanded search facilities (BT)
• different views on semantic structure of the
  document (AIdministrator)

29
SwissLife, CH
30
Expected achievements / impact

• A methodology for introducing knowledge
  management concepts and tools into enterprises
• An intelligent search tool and a tool environment
  for maintenance, conversion, and acquisition of
  information sources
• Three large case studies help developing the
  technology according to actual needs of large
  and/or virtual organisations

31
Useful Links
IST Homepage http//www.cordis.lu/ist/ On-To-Know
ledge Homepage http//www.ontoknowledge.org/ OIL
Homepage http//www.ontoknowledge.org/oil/ Semant
icWeb.org Homepage http//semanticweb.org