Inferring user Search Goals with Feedback Sessions using STC

The information explosion on the internet makes high demand on search engine. But the existing search engines are not satisfying the user-s needs. Most of the search engines returns irrelevant document in response to the user search query. Sometimes the search query may be imperfect description. For such imperfect query the search engine cannot retrieve relevant document to the user-s needs inferring user search goal can improve search engine relevance and user search experience. In this paper propose a new method to infer user search goal by analysing search engine query logs. For better performance construct a feedback session from click through logs of search engine and map these sessions in to corresponding pseudo documents for clustering. By using clusters organize the search results according to the user relevance. By using CAP algorithm optimize the number of clusters and evaluate the performance of inferring user search goal.