An Extension to New Algorithm for Feedback Session

Internet is widely used by users to satisfy various information needs. This paper proposed a novel approach to infer user goals. Pseudo documents are generated through feedback sessions. We introduce new criterion”classified average precision(CAP)”, for evaluate performance of inferring user search goals. Results are represented on search engine to validate the effectiveness of our work. Extraction of interesting information from web data has become more popular and result of that web mining attracted lot of attention in recent times.