Semantic Indexing and Fuzzy Relevance Model in Information Retrieval

Let us suppose that an information retrieval system comprehends the semantic content of documents and reflects the preferences of users. Then the system can more effectively search for the information better on the Internet, and may improve the retrieval performance. Therefore, in the present study, a new information retrieval system is proposed by combining a semantic based indexing and a fuzzy relevance model. In addition to the statistical approach, we propose the semantic approach in indexing based on lexical chains. The indexing extracts the semantic concepts in a given document. Furthermore, a fuzzy relevance model combined with the semantic index calculates the exact degrees of relevance between documents based on the user preference. The combination of these notions is expected to improve the information retrieval performance.

Authors and Affiliations

1. Department of Computer Engineering, Kyungpook National University, Sangyuk-dong, Puk-gu, Daegu, Korea

   Bo-Yeong Kang & Sang-Jo Lee

2. Department of Computer Science, Korea Advanced Institute of Scienceand Technology, Guseong-dong, Yuseong-gu, Daejeon, Korea

   Dae-Won Kim

Reprints and permissions

Policies and ethics