File Search with Query Expansion in a Network System(s)

Abstract

The amount of information in the Internet is growing fast; Searching for information has become an important issue; however the user queries impact the effectiveness of retrieving information that users need. The objective of Query Expansion is to find additional and more relevant results. This article used different similarity measures (Cosine, Jaccard, Dice similarity functions) in VSM on three Genetic Algorithm approaches, each similarity function used as fitness function, one point crossover and new selection method based on rank selection is used. The NSC (National Science Council, Taiwan) document data collection is used in this study. Our results show that QE methods increase the precision rates and the recall rates of information retrieval for dealing with document retrieval . Also we present a network system that consists of many servers to decrease the amount of workload from Main server.

Keywords: information retrieval, vector space model, similarity measures, genetic algorithm, query expansion.