The Design And Implementation Of Personnalized News Fetching And Polymerization System

The number of Chinese Internet user andwebsites has been growing rapidly with the development of the Internet and the rise of mobile Internet technology, and the influenceof networks are expanding. Portals, social networks and micro-blog contains lots of dynamical news and information, becoming an important source for the latest information.So, in the face of such a flood of news information, how to design a personalized news polymerization system to push the news and customized content to users timely and dynamically, so as to save the searching time, becomes a hot area of current research.The main subject of this thesis is studying and designing personalized news fetching and polymerization system to provide users with a personalized news service. The system takes advantage of the network information crawling, web-based text mining and personalized recommendations technologies. Crawling network information using web crawlers technology and this area is already quite mature, besides, there are many open source tools, such as Nutch, Larbin, Heritrix, WebLench, etc. In this study, we use Nutch,Then,via text clustering method and topic detection technology,we aggregate and analyze information on the network to find the hot news.By analyzing the historical behavior of users, we make interest model for each user, as a basis to predict users’s interested news and make some personalized recommendations.Eventually, with the realization of a B/S architecture personalized news recommendation system, users can see the recommended news content with their mobile client or web client. The work and innovation of this thesis are:(1) Implementing the cluster and analysis of news in use of TD*IDF algorithm;(2) Implementing the personalized system for the interest of users;(3) Implementing a web-based visualization system.Based on users’interests and behavior features, personalized news recommendation system can help users easily access the news information they are interested in from the mass of information on the Internet, explore the contents that users may be interested in, besides, it does not require much more time for users to search information and helps achieve a win-win solution between the benefits of newswebsites and users.