Christian Halaschek-Wiener (Department of Computer Science, The University of Maryland)
James Hendler (Department of Computer ScienceThe University of Maryland)
Abstract:
Syndication systems on the Web have attracted vast amounts of attention in recent years. As technologies have emerged and matured, there has been a transition to more expressive syndication approaches; that is subscribers and publishers are provided with more expressive means of describing their interests and published content, enabling more accurate information filtering. In this paper, we formalize a syndication architecture that utilizes expressive Web ontologies and logic-based reasoning for selective content dissemination. This provides finer grained control for filtering and automated reasoning for discovering implicit subscription matches, both of which are not achievable in less expressive approaches. We then address one of the main limitations with such a syndication approach, namely matching newly published information with subscription requests in an efficient and practical manner. To this end, we investigate continuous query answering for a large subset of the Web Ontology Language (OWL); specifically, we formally define continuous queries (i.e., subscriptions) for OWL knowledge bases and present a novel algorithm for continuous query answering in a large subset of this language. Lastly, an evaluation of the query approach is shown, demonstrating its effectiveness for syndication purposes.
Slot:
New Brunswick, Friday, May 11, 2007, 10:30am to 12 noon.