Design And Implementation Of News Intelligent Recommendation System Of Tencent Mobile Website

With the development of the Internet, information overload has become an urgent problem that need to solve on the Internet area, and the ways to solve this problem appear more and more. In addition to the information classification and provide a better search engine, you need personalized intelligent recommendation system to get the information which can not be described. Personalized intelligent recommendation system is a project which provide users with personalized intelligent recommendation function to guide users to read or browse the product and information with purpose by using data mining technology. As a result, the Internet has transformed itself from a stage of "information waiting for query" to a stage of "information confirming delivery".In order to solve this problem, I decided to research this direction as my topic, to explore how to guess the user’s interest and push their favorite articles more precisely. This system relies on the mobile phone tencent, and realized the news intelligent recommendation system with the ordinary information page existing at the same time.Mobile phone tencent news intelligent recommendation system mainly divided into the theme management module, recommend module and display module. System divided into 71 theme according to source and type of the article. Theme management module is only designed for our news editor, and they can search, deal with, delete or download pos information combining various conditions. News recommendation is an important part of this system and it mainly implement three funtions which include recommend algorithms of shopping cart, building of user model, and integration scenarios. In the recommended module, the system uses the FP-Growth shopping cart algorithm to analyze the history of the user’s dataand get the frequent sets based on the article topic. By collecting the user’s data, accumulating and calculating with the user’s interest associating with news topic collection, eventually we make users more accurately read their own articles. Display module is the entry of the recommendation system. Users can browse the news content which be recommended in display interface and record their own behavior in the display module.In terms of test, I have finish functional and non-functional testing for Mobile phone tencent news intelligent recommendation system, and the results can meet the expected requirement, and can be used for news editor and ordinary users. The whole system has been used for half a year, and we have many good user feedback.