Inferring User Search Goals using Feedback

For a wide subject and ambiguous query, diverse users may have distinctive search objectives when they submit it to a web search engine. The derivation and examination of user pursuit objectives i.e. goals can be extremely helpful in enhancing web search tool significance and user experience. In this, we propose a methodology to gather user look objectives by examining internet searcher query logs. In the first place, we propose a structure to find diverse user hunt objectives down a query by grouping i.e clustering the proposed feed-back sessions. Input sessions are developed from user navigate logs and can effectively reflect the data needs of users. Second, we propose a methodology to produce pseudo documents to better represent the feed-back sessions for grouping. At last, we propose another measure "Classified Average Precision (CAP)" to assess the execution of deriving user look objectives.