Tuning topical queries for effective information retrieval

Abstract

How anyone can find the desired bit of information with respect to his/her own context from the ocean of information resided in multiple databases and text repositories growing at an enormous rate. Information retrieval systems (IRS) are use to find information as output with respect to the user query as input. Effectiveness of information retrieval system hugely depends upon the query formation. Various factors affecting query formation are media expertise, domain expertise and type of search [1,2]. Search engines are necessary tools for information retrieval from World Wide Web. Conventional search engines like Google, Yahoo etc. have huge amount of data. To retrieve the information from these conventional search engines which serve the population on the whole without much concerning about user context require user query expressive enough about user context and need. In our paper we have proposed a model to build a context based search engine on the conventional search engine using genetic algorithm. We have tried to find out good query terms in context of user to find user specific retrieval. We have used these terms for query expansion or reformulation.