IST

1
Information Society Technologies Programme
From content processing to knowledge
management Pierre-Paul Sondag European
Commission DG INFormation SOciety
/D5 Information Access Filtering and
Handlinghttp//www.cordis.lu/ist/ka3/iaf/index.ht
mpierre-paul.sondag_at_cec.eu.inttél. 352 / 4301
34480
2
Exploiting the full potential

• new data capture and sensor technologies will
  generate Pbytes
• new forms of content will represent complex or
  real life objects
• increasingly accessible from mobile platforms
• the full potential can only be reached by making
  content machine understandable
• in particular, the World Wide Web is an
  exponentially increasing amount of flashy data,
  making a vast information space, but only and
  barely exploitable by humans

3
Adding structure to metadata
- Metadata are enriched with structured
taxonomies called ontologies
- Ontologies have attributes that can be
processed automatically
- Ontologies enable inferences
?
- Ontologies should be as far as possible machine
generated
- But users/developers should keep the overview
of Ontologies
4
Call 7 on Semantic Web Technologies

• The notion of Semantic Web was first coined by
  the W3C
• def. approach to reduce the burden by the user
  and to automate the processing of Web content
  making it machine-understandable rather than just
  machine-presentable
• adding semantics (metadata ?) to the web
• to link separate applications to shared
  ontologies
• more technology but simpler for the user

5
Components of the Semantic Web

• knowledge representation systems and technologies
• using standards to enable interoperability, XML,
  RDF, SOAP, OIL
• multimedia dimension
• software agents
• content analysis

XML
SOAP
RDF
OIL
6
Call 7 Semantic Web Technologies
Focus on

• methods and tools for coding and structuring
  content
• derivation of semantic attributes of content in
  particular images, audio, videofeature
  detection, video segmentation, post-processing
• tools for knowledge discovery, intelligent
  filtering, profiling, information agents, query
  languages
• information visualisation, making semantic
  structures palpable to the end-users

7
definition of aims within FP6

• Policies for the knowledge society were agreed on
  theLisbon council 2000 and Stockholm summit
  2001DG INFSO organised knowledge technologies
  workshopApril 2001, in Brussels
• to highlight the importance of the subject
• to examine the state of the market and of
  research
• to identify key challenges
• to provide input to FP6 programme

27 leading experts from industry academia in
content technology identified following key
issues
8
Usability

• Key issue 1 usability
• creating value is not enough, we need to express
  those values in terms of enjoyment and enrichment
  for the user
• usability turns content into knowledge
• extensive metadata are not enough it is necessary
  to model the discovery process
• learning from user behaviour should benefit to
  the knowledge representation process

9
Content as knowledge

• Key issue 2 content as knowledge
• multimedia nature of the challenge, we are
  entering the age of the sensors
• trustability and certification of content
• versioning and multilinguality, helping every
  group of users to keep control of its viewpoint
• preservation of the collective memory of a
  community

10
Interoperability

• Key issue 3 interoperability and standards
• challenge of linking ontologies in different
  sectors
• bottom-up approach, reflecting user interaction
  with knowledge
• to cope with the volume of content, metadata have
  to be increasingly machine generated
• same standards for interoperability to be used
  for the content and the metadata
• confusing generation of new standards in discrete
  areas of knowledge representation

11
Communities

• Key issue 4 communities and portals
• diversity of access mode, ambient intelligence,
  mobile, unifying different views on ontologies
• advanced visualisation features, location
  sensitivea picture is thousand words worth
• access to collective memory, intelligent
  portalsit is important to know what the
  community know
• filtering, personalisation, transformation of
  content filtering to enable knowledge

12
knowledge repositories

• Objective 1 knowledge representation is based on
  integrated repositories, requirements include
• content based search engines
• content based indexing, feature extraction
• metadata generation driven by users views and
  behaviour
• methods to define users profiles to support
  collaborative agents
• dynamic content with user-interaction

13
knowledge environment

• Objective 2 creating an enjoyable access by
  interactive visualisation of real-world or
  abstract metaphors
• multi-user visualisation framework
• visualisation of retrieval results
• query languages usable without knowing the
  structure
• integration of NLP
• quality and trustability
• scalability
• security

14
Call 8 bridging to FP6

• Develop research roadmaps for FP6 instruments and
  build constituencies
• Target areas include
• Knowledge and interface technologies,
• Dynamic interactive content etc
• AM or TN public-private partnerships
  European-wide consensus building
• Also pioneering RTD in above areas
  dissemination.

15
FP6 A new structure
16
key issues in FP6

• More focussed, strategic themes
• New instruments
• integrated projects
• networks of excellence
• Also continuity normal RTD projects, stairs
  to excellence.
• IST budget 3.9 M ?
• Decision - 2nd half 2002

17
knowledge technologies
enables access to content
enables content handling automation
increases the value of content
...and hence turns it into knowledge
18
Conclusion
knowledge is becoming the foundation of economic
and social relationship
it is the key to economic competitiveness
and can also provide better quality of life