KAAS
From Roadmap
Contents |
Knowledge Acquisition, Annotation and Sharing
Objectives ad Scope
This service will cover knowledge acquisition and knowledge sharing in semantically rich environments. Acquisition is defined here as the process of:
- annotating existing multi and cross-media resources
- unstructured: e.g. texts, images, data, videos, etc.; annotation can be ontology-based (as in OntoMat and AktiveMedia) or free-text (as in Annotea) or folksonomy-based (as in Web 2.0);
- structured: e.g. existing databases (as in the Web of data); annotation will be here ontology-based;
- information and knowledge integration across resources and media.
Sharing is here the process of making the knowledge available independently from source or media at the right time in a form suitable to a specific user. The area remit will involve Human Computer Interaction, Social Science, etc.
Acquisition and sharing enable the effective exploitation of core semantic web technologies (such as reasoning, representation and web services) into applications. The service will constitute a focus point for discussion and research around the main challenges of acquisition and sharing.
Context
A number of challenges are emerging for knowledge acquisition and sharing in Semantic Web scenarios, the major one being its large scale. Such scale emerges along a number of dimensions, from the sheer size of ontologies (which for some domains can be very large), to the number of instances (billions of triples), to the number of documents (millions to potentially billions of multi-media documents), to millions of users. Also the recently emerged Web 2.0 is posing new challenges to the Semantic Web towards:
- new ways of annotation (e.g. folksonomies); the relation between tagging and ontology-based annotation is largely unexplored and is becoming a major research issue.
- the social use of information, which adds a new dimension of knowledge to the personal and organizational knowledge traditionally addressed by our community.
Concerning Knowledge acquisition, a preliminary list of challenges can be listed:
- Ability to scale to large amount of data and large ontologies
- Ability to acquire knowledge within and across media
- Ability to integrate and fuse knowledge across archives and collections of documents.
Challenges for knowledge sharing and reuse are:
- Ability to focus searches within very large archives and with large ontologies
- Creation of an infrastructure able to share across large amount of documents and knowledge
- Ability to reason over large amount of knowledge
Moreover some challenges can be identified across acquisition and sharing, i.e.:
- Contextual design of acquisition and sharing technologies so to overcome some of the limitations of acquisition technologies
- Empowering reuse information while knowledge is created
Approach
According to the strategy adopted for the roadmapping service, the goal of the activity is to identify the final list of challenges and to derive a timeline for development.
Outcome
The outcome will be a roadmap for Area KAAS.
Roadmapping: Associated Project
The roadmapping exercise will be carried out in cooperation with the X-Media project consortium who are planning a roadmapping exercise for June 2008 Other projects are being invited to join.
Roadmapping: who
The following people are proposed to participate in the roadmapping exercise and workshops (preliminary list)
- Enrico Motta, OU
- Martin Dzbor, OU
- Nigel Shadbolt, U Southampton
- Steffen Staab, U Koblenz
- Diana Maynard, U Sheffield
- A Bernstein, U Zurich
- Y Sure, SAP
- A Lavelli, FBK Irst
- J Iria, U Sheffield
- Andrew McCallum, U Massachusetts
- Y Kompatsiaris, U Thessaloniki
- G Satta, U Padova
- L Schmiedt-Thieme U Hildesheim
- L Gilardoni, Quinary
- J Throssell Rolls-Royce
(...)
Roadmappping Workshop: when
The expected period for the organisation of the roadmapping workshop is between June 2008 and November 2008