Intelligent Database Systems Lab 國立雲林科技大學 National Yunlin University of Science and Technology 1 An Adaptation of the Vector-Space Model for Ontology-Based Information Retrieval Author: PabloCastells, MiriamFernandez and DavidVallet, Reporter: Wen-Cheng Tsai 2007/02/13 TKDE,2007
Intelligent Database Systems Lab N.Y.U.S.T. I. M. 2 Outline Motivation Objective Method Experience Conclusion Personal Comments
Intelligent Database Systems Lab N.Y.U.S.T. I. M. 3 Motivation Semantic search has been one of motivations of Semantic Web since it was envisioned.
Intelligent Database Systems Lab N.Y.U.S.T. I. M. 4 Objective We propose a model for the exploitation of ontology-based knowledge based to improve search over large document repositories.
Intelligent Database Systems Lab N.Y.U.S.T. I. M. 5 Method Arts Irises Arts painting instance Van Gogh Weighting annotations Ex : Ranking algorithm Example “Players from USA playing in basketball teams of Catalonia”
Intelligent Database Systems Lab N.Y.U.S.T. I. M. 6 Root ontology classes Fig. Root ontology classes 1. DomainConcept 2. Document 3. Topic
Intelligent Database Systems Lab N.Y.U.S.T. I. M. 7 Example Player Bramlett Derrick Alston John
Intelligent Database Systems Lab N.Y.U.S.T. I. M. 8 Experience a b c (a) ”News about banks that trade on NASDAK, with fiscal net income grater than two billion dollars.” (b) ”News about telecom companies.” (c) ”News about insurance companies in USA.”
Intelligent Database Systems Lab N.Y.U.S.T. I. M. 9 Conclusion Our approach can be seen as an evolution of the classic vector-space model, where keyword-based indices are replaced by an ontology- based KB, and a semiautomatic document and weighting procedure is the equivalent of the keyword extraction and indexing process. We show that it is possible to develop a consistent ranking algorithm on this basis, yielding measurable improvements with respect to keyword-based search.
Intelligent Database Systems Lab N.Y.U.S.T. I. M. 10 Personal Comments Advantages ─ Improve conventional keyword-based search Disadvantage ─ … Application ─ Information retrieval