基于协同过滤和应用风险的移动应用推荐系统

With the development of mobile Internet, the number of mobile applications(App) increases tremendously, making it much more difficult for users to choose the Apps they interested, and there are usually some Apps which have the risk of personal information leakage. For these two issues, this paper proposes a mobile app recommendation system schedule which based on collaborative filtering and application risk. The schedule firstly produces the primary recommendation results by the improving collaborative filtering algorithm, then evaluates the applications according to application's permissions, acquired the application risk value, combined with these two metrics, the system could recommend a proper app for users. This paper summarized the design of the system and its implementing, the experiment shows that the improved algorithm raised the accuracy of recommendation result. Experimental results show that the risk assessment method for the application of this paper is effective. Finally, the paper analyzed the ways to optimize this recommendation system.