Personalized Online News Service Of User Interest In The Learning Algorithm

The Internet news, with its special character of timeliness, is now spreading into the daily lives of users. Compared with the traditional news, Internet news has more advantages, such as being browsed by readers in the first time, having high volume and plenty of information, having diversified format, easy to be stored and copied, etc. However, there still exists common "rich data, poor information" phenomenon of content service in Internet news service. So, the Internet news has to take the personalized reading satisfaction of readers as the first aim to gain a good future. The HTML technology and HTTP protocol, which the Internet news service usually used, cannot get rid of the push-out service of traditional news medium and there is no common definition of personalized need of readers in service system.First, this article surveyed the Internet news, and point out its character and shortage; summarized user's habit when he reading the Internet news, and educe the user's reading character; summarized personalized service technology. Second, this article summarized the prophase work of the lab, designed the news reader's personalized customization and reading news-service architecture and the Internet news service model architecture.A new technique of create user's interest model and the user's interest model's updating algorithm are designed based on the forward summarization and the prophase work of the lab. It increases the precision of the personalized news service effectively; avert the information's repetition and deflection. And at the same time, it improved the shortage of the prophase work of the lab.Finally, a personalized Internet news service system that based on Web service and B/S architecture was implemented. The new technique of create user's interest model and the updating algorithm of user's interest model were used in the system. And it proved the algorithm's feasibility.