The Design And Implementation Of News Recommendation System Based On Android

With the popularity of smartphones, the entrance of the internet split from web portals on traditional PC top to mobile internet gradually, in the sphere of news reading, people are more dependent on the news client application on mobile device. However, the explosive of internet data make users find it’s difficult to achieve interested news from mass news. Fortunately, personalized recommendation technology is applied to news reading application on mobile phones, with the purpose to set users free from mass information. It could accurately help users access to satisfied interested news rapidly, what is a worthwhile direction to explore in.In this thesis, we design and implement an Android application for news recommendation system. Firstly, we analyze the business requirements and design the functional modules of the system; meanwhile, a UserCF(User Collaboration Filtering) based recommender technology is designed, what is compatible to the feature of news, it collect user data for calculating the reading time-consuming factor to modify users’ preference value, and consider hotspot of news to punish the user similarity value; Then, we rank the news from user’s neighbors set by Top-N method, what comes as recommender news list; Lastly, after completed preliminary design and algorithm of system, an android application is developed to display the recommended news, also we construct a SSH(Spring+Struts2+Hibernate) framework based news administration background to realize the recommender by Mahout components, with manage function for news and users.The prototype of the system has been completed; it has tested to reach the design expectation. The system handle the users’ simple operation data from the Android client, to update the user interest model, to recommend news quickly and precisely; Moreover, the system innovatively apply a User Collaborative filtering based algorithm to news recommendation.