Extracting Query Expansion Terms Based on User´s Search Behavior

Search engine(SE) can obtains rich information from the Internet as a powerful tool, but most of them are not relevant to the searcher´s intended requirements. The main factor is query terms can not depict the user´s personality with the ambiguity of the language. To solve this problem, a method of extracting query expansion terms based on user´s behavior is proposed in this paper. The method analyzes their potential relevance of the user´s search history and click history. The search return sequence and the user clicks sequence are considered as well. So we can extract terms which can depict the user´s requirements. An experiment illustrates that our method can effectively extract the relevant terms and improves the quality.