Web Search Result Clustering using Heuristic Search and Latent Semantic Indexing

Giving user a simple and uncomplicated web search result representation is an active area of Information Retrieval research. Traditional search engines use the hyperlink structure of the web to retrieve documents or pages and give them in a ranked fashion to the user. In this paper, we propose a technique for grouping web search results into meaningful clusters. The proposed method performs heuristic search on the query result graph to prune undesired edges to form cluster and carries out Latent Semantic Indexing within these clusters to make them refined, meaningful, and relevant to the query.