Personalized User-Query Semantic Clustering Using Search Click Information

Because of query's semantic ambiguity, search process of general SE can not meet the personalized demand of users concerning personal interests and professional backgrounds. To resolve this problem, a new personalized user-query semantic clustering approach is proposed in this paper. The search engine user logs are valuable resources which obtain the rich history information of user access records which reflect the user's interests and domain knowledge. For every specific user, we get three semantic relationships between user-query and their search click information, such as query contents, click sequence and selected documents. In this way, user-query semantic similarity can be calculated using search click information, then user-query can be clustered and disambiguated based on user's interests. Through the personalized query clustering to guide topic crawling, you can concentrate on more in-depth in the user's interesting field.