Research On Mobile Search Algorithm Based On User Context Associated With The Interest Model

With the continuous development of the mobile Internet and mobile intelligent terminals, end users are getting more and more dependent on the use of the mobile terminals for information acquisition, and mobile search application has become one of the most frequently used applications by mobile Internet users. Nowadays, how to provide search results that match end users interest has become one of the hot research spots.For mobile search application, the characteristics of mobile terminals and search application were fully considered in this thesis, and the methods for constructing and updating the users’ interest model were discussed in details. In the research, a dynamic user interest model based on the theme concept hierarchy and vector space model was created, which combined the current mobile environment with users’ interest information. Besides, a updated method of users’ interest model was proposed which was according to the similarity of situations and combining long and short term interest feature.On this basis, the study and research on how to use the situations associated with interest model to influence the final search results and improve user satisfaction with mobile search results was conducted in-depth, two algorithms which were query expansion algorithm based on clustering analysis of users interest model and improved rank algorithm based on the traditional PageRank algorithm were proposed, the improved rank algorithm adds users interest in the mobile environment to the process of ranking and corrected the algorithm by users feedback on search results.At last, the average precision and nDCG evaluation system were used to evaluate the model and algorithms, experiment results show that the proposed model and algorithms can accurately describe users’ interest and its dynamic process in mobile environment, the ranking result compared to the traditional PageRank algorithm has higher precision and nDCG values.