基于大数据的新闻媒体预测推荐系统报告

Abstract

The site now offers the same content to all users. Users can only refer to relevant navigation and leaderboards, and use search engines to find what they need. Website recommendations to users of information is only some of the hottest news, newspapers, e-books, and can not be based on the user’s own interests need for personalized recommendation services. This form of recommendation can become a non-personalized recommendation service, it has many disadvantages. Therefore, personalized news recommendations are a solution to this problem. By building a user interest model, tracking and recording the browsing and clicking behavior of different users, and then using this information to update the user interest model, based on the analysis of the data in the model, we can predict what the user would be interested in and what might be interested in.