Emergent Semantics and Ontology Evolution
|Start||November 12 2007 9:00 (iCal)|
|End||November 12 2007 17:30|
|Papers due:||August 14 2007|
|Submissions due:||August 14 2007|
|Notification:||August 30 2007|
|Camera ready due:||September 15 2007|
Event in series ESOE
The Semantic Web and collaborative tagging are two complementary approaches aiming at making information search, retrieval, navigation and knowledge discovery easier. While the Semantic Web enforces semantics top-down via the use of ontologies, collaborative tagging tries to obtain semantics in a bottom-up fashion. Del.icio.us and flickr are success stories of collaborative tagging; the winners of the Semantic Web Challenge demonstrate the success of the Semantic Web. Still, both approaches face open issues. For the Semantic Web, ontology engineering, in particular, large-scale ontology construction, has been a bottleneck. While effort and progress have been made in ontology matching, alignment, versioning and learning, it has become clear that constructing large ontologies requires collaboration among multiple individuals or groups with expertise in specific areas. Also critical is the ontology evolution in the open, dynamic Web environment in order to keep pace with the Web dynamics. For collaborative tagging, tags (metadata) can be generated in large-scale and capture users’ collective wisdom. However, large-scale tagging usually degrades the performance of re-findability due to the ambiguity of uncontrolled vocabulary and the flat structure of “tag soup”. In such a case tagging alone is not helpful at all for solving the problem. Bundles, classification, relations or tagging of tags are some promising ways to enforce some kinds of structure for tags in order to enable scalability and findability.
There is a growing interest in marrying the two paradigms in order to create large-scale semantic and intelligent content. The basic idea is to 1) derive emergent semantics from community-based collaborative interaction as demonstrated by Web 2.0 application, in particular, folksonomic tagging; 2) extract and formally model emergent semantics in structures, such as ontologies; 3) construct and evolve ontologies as emergent semantics from collaborative applications are of dynamic nature; and 4) enhance collaborative applications with formal ontological structures. Against this background, the proposed workshop aims to bring together researchers and practitioners in the relevant fields of the Semantic Web, ontology engineering, folksonomy, social Web, artificial intelligence, machine learning, information integration and relevant application areas (e.g., bioinformatics, enterprise knowledge management, e-science, e-government, medical informatics, social informatics, among others) to discuss the current state of the art and open research problems in emergent semantics and ontology evolution. A secondary goal of the workshop is to facilitate collaborations between different research groups.