An Innovative Method for Concluding User Search Intention Using Feedback Session

Abstract

With the advent of computers, it became possible to store large amounts of information and finding useful information from such collections became a necessity of today’s world. For a broad-topic and ambiguous query, different users may have different search goals. The inference and analysis of user search goals can be very useful in improving search engine relevance and user experience. Hence, in this paper we have tried to improve the efficiency of the search engine by combining inferring user search goal with feedback sessions. Initially, we composed a framework to implement different user search goals for an ambiguous query with the help of clustering the proposed feedback sessions. The user needs are reflected efficiently through the feedback sessions built by the user click-through logs. Second, to represent better feedback sessions for clustering we generate the pseudo document. Finally, we proposed a new criterion “Classified Average Precision (CAP) to evaluate the performance of inferring user search goals. Therefore, when users submit their queries, the search engine can return the results that are categorized into different groups according to user search goals online. Thus, users can find what they want conveniently