| With the rapid development of the mobile Internet, more and more mobile application developers are in a fierce competition in the mobile applications market. Faced with a large number of mobile applications, the users on the mobile Internet need to spend a lot of time to select the application they are interested, the mobile application developers also need to spend more money to get a new user who would use their mobile applications. To solve this problem and help more applications could be showed to the user on the mobile Internet, the personalized recommendation system for mobile applications is proposed.Based on the traditional recommendation system and combined with the characteristics of the mobile Internet, this thesis designed and implemented a personalized recommendation system for the mobile applications. First, this thesis introduced the background and the content of the subject and the knowledge related to the personalized recommendation system for the mobile applications. Secondly, this thesis analyzed the requirement of the user and the application developers on the mobile Internet and proposed a way to help the application developers to recommend their mobile applications each other. Then, according to the habits of the user on the mobile Internet, this thesis proposed a new algorithm which was based on the implicit feedback of the users and improved the weighted Slope One recommendation algorithm, and do some experiments to prove the new algorithm was more accurate. This thesis also proposed a solution which was more suitable on the mobile Internet for tracking a user. According to what has been mentioned above, this thesis designed a personalized recommendation system for the mobile applications and implemented the recommendation module, the logging module and the offline-predict module. Finally, this thesis introduced how to make a test and experiment to prove that the system was correct, available and scalable. |
PDF Full Text Request