Integrating Relation and Keyword Matching in Information Retrieval
We propose an information retrieval (IR) model that combines relation and keyword matching. The model relies on a novel algorithm for relation matching. The algorithm takes the advantage of any existing relational similarity between document and query to improve retrieval effectiveness. If query concepts(terms) appearing in a document exhibit similar relationship then the proposed similarity measure will give high rank to the document as compared to those in which query terms exhibit different relationship. A conceptual graph (CG) representation has been used to capture relationship between concepts. In order to keep the approach computationally simple a simplified form of CG matching has been used instead of graph derivation. Structural variations have been captured during matching through simple heuristics. CG similarity measure proposed by us is simple, flexible and scalable and can find application in many related tasks like information filtering, question answering, document summarization etc.
KeywordsInformation Retrieval Retrieval Model Vector Space Model Conceptual Graph Concept Node
Unable to display preview. Download preview PDF.
- 2.Chein, M., et Mugnier, M.-L.: Conceptual graphs are also graph. Research Report, 95-004, LIRMM (1995)Google Scholar
- 4.Khoo, Christpoher, Soo-Guan: The Use of Relation matching in Information Retrieval. LIBRES: Library and Information Research 7, 2 (1997)Google Scholar
- 5.Liddy, E.D., Myaeng, S.H.: DR-LINK: a system update for TREC-2. In: Second Text REtrieval Conference (TREC-2) (NIST-SP 500-215) NIST, Washington, DC, USA, pp. 85–99 (1994)Google Scholar
- 13.Smeaton, A.F., O’Donnell, R., Kelledy, F.: Indexing structures derived from syntax in TREC-3: System description. In: Harman, D.K. (ed.) Overview of the Third Text REtrieval Conference (TREC-3) NIST SP-500-225, pp. 55–67. NIST, Gaithersburg (1995)Google Scholar
- 15.Sparck Jones, K.: Summary performance comparisons TREC-2, TREC-3,TREC-4, TREC-5. In: TREC-5 Proceedings (1997), http://wwwnlpir.nist.gov/TREC/trec5.papers/sparckjones.ps
- 16.Xu, Y., Benaroch, M.: Information retrieval with a hybrid automatic query expansion and data fusion procedure. Information Retrieval 7, 1–25 (2004)Google Scholar