Efficient Query Suggestion System using User Search Log

Abstract

With the growing information burst on the World Wide Web, internet has placed high demands on search engines. Existing search engines provide most of the features for user query. But users of internet are not satisfied with them as they return thousands of documents in response to user query. So to develop user search intent application is challenging, satisfying increased expectations & diverse needs of user. Recorded user search logs are analyzed and used to form clusters. This clustering of user query is imperative for filtering the relevant results, so in the proposed system first approach mines frequent query patterns from users search history using FP growth, if user want to select any query from his/her previous search history then he can. Otherwise he will enter new query and second approach identifies clusters of queries from all users search history those clusters of queries which is similar to current query are query suggestions. Thus, by automating the optimization process of searching on web; we can minimize user efforts; maximize user satisfaction for getting desired search.