Authors : R.M. Periakaruppan and R. Nadarajan
Abstract: Keyword based Information Retrieval (IR) Systems fails when there is no exact matching. Hence, IR Systems more focused on user relevant information retrieval. In this study, researchers proposed a technique to improve the searching based on Fuzzy set similarity measure, domain knowledge representation and semantics. In the proposed method, researchers extract candidate keywords from each document which reflects the topic of the document and map them with a suitable domain knowledge classification system. The mapping process associate a semantic weight to each keyword based on Latent Semantic Indexing (LSI) which reflects the significance of each keyword in the document with respect to a domain. As a result of mapping process the keywords along with their semantic weights are represented as XML document and they are clustered based on similarity measure. The experiments shows that the proposed similarity measure yields better results when compared with conventional similarity measure techniques.
R.M. Periakaruppan and R. Nadarajan, 2013. An Improved Information Retrieval System Using Fuzzy Set Similarity Measure. Asian Journal of Information Technology, 12: 45-52.