Personalized Query Auto-Completion With Time-Sensitive Analysis

Internet information is huge and complex currently, so it is importante for search engines giving internet users a good information searching experience. Query auto-completion of search engines can provide a better experience for the users, so query auto-completion technology has applied in search engines widely. But most of the technologies rely on wide-wisdom. With user’s search intention changing over time, time-sensitive query auto-completion (TSQC) comes to be a hot area of research.Different from traditional query auto-completion, recommendation list of TSQC is according to the attaching time features of the query words. Because there is no auto-completion algorithm focusing on both personality and time-sensitive, we developed a method called personalized time-sensitive query auto-completion (PTSQC) to recommend queries. First, we put user’s operation data on his/her own computer. Then, the dependency between user’s needs and timeline is described by Bayes’ rule and time-series forecasting models. At the end, the queries are sorting based on weight score. At the end of paper, the comparison between PTSQC and popular method is given. The experiment results show that PTSQC performs better than the other.