Conceptual Query Expansion Model for Web Information Retrieval

The process of retrieving relevant documents from user query is to begin with the clustering of documents with high semantic similarities between terms, and lower inner noise values. Here, the research extends normal keywords document clustering techniques in automatic thesaurus construction to building a Concept Based Thesaurus Network. The applied concept matching algorithm uses the Multi-Fuzzy Concept Network to generate sub clustered documents with relative degree of relationship across the clustered document. The proposed system achieved a higher cohesion rate between concepts and lower entropy rate in document. Also, a concise and relevant potential retrieved document were better ranked when compared with other existing document clustering techniques.