Ontology Summit 2008/Work Session

From Tetherless World Wiki

Event Info [ Edit ]
Type	Talk
Title	Work Session
Location
When	2008/04/29 - 2008/04/29
Abstract	review talks presented yesterday, and discuss several highlights
Webpage
More Details (out-links) Part of Ontology Summit 2008 Speaker(s) Leo Obrst, Mark A. Musen What Links Here (in-links)

recap yesterday

Frank's talk

• why OOR is needed?
• ontology search
• graph representation based approach (semantic network is the fundation)
• partial orders, modularize and relate ontology
• provenance, metadata for ontologies
• other tools, version, import/export
• SOA services, REST is preferred

architecture

• archival approach
• federated approach
• instance data of ontology
• metadata of ontology
• representation of ontology
• mappings and integration

quality and gatekeepting

• the two concepts are related, so we should avoid duplication
• quality, trust
• governance, promote better ontology, copyright

ontology of ontology

• since last summit: we moved from upper ontology and mapping ontology to "ontology metadata" and OOR

discussions

• we need a test case with 7 ontologies to initiate our investigation of OOR
• mapping ontology: it is important but not required
• expectation of ontology quality
  • (Ravi) we need to avoid junks with minimal gatekeeper
  • (Steve,Bruce) submit free, view with filter, e.g. social-community filter, (berry) central-expert filter
• (frank) editing ontology need identity management

break down discussion

• (elisa) W3C deploy work indicuate that well used RDF vocabularies: (i) does not have enough documentations on itself, change management, (ii) does not have policies for how to contribute, revision.
• ontology users are under represented