Research And Implementation On Blogger Personalized Modeling

In 2002, the concept of Blog was introduced to china and was developing very quickly. Untill 2006, Blog has received a very huge scale. The amount of Blogs registered is more than 33 million. "Blog" has become one of the most popular applications on the internet. However, there are only 44% among all the Chinese Bloggers that regularly update their Blogs. So there is a big waste in space. Moreover, Bloggers often switch between different Blog Service Providers. The reason on earth is that the services provided by the BSP are all the same and are boring. The services cannot reach Bloggers' satisfaction. Therefore, how to improve the services is very important to BSP.Viewing from the content of Blog, the Blog articles reflect the personalized points of the Bloggers. Using this advantage, we can analyze the content of Blog articles and establish a personalized model for every Blogger to understand his interests and requirements. Then we can provide personalized services based on the Blogger model to satisfy Blogger' requirements. This method will probably advance Bloggers' satisfaction and loyalty and will also enlarge the scale of Bloggers.This paper has put forward and implemented a personalized Blogger modeling method that establishes a long interest model and a short interest model for every blogger. The long interest model reflects things in which the blogger always interests and will decline gradually. And the short interest model records the blogger's recent interests. If the blogger cares about something continuously in a period of time, it will be added into the long interest model.The emphasis of the work is to ananlyze the personalized Blogger modeling with two examples of application. The reasons are below: on one hand, the personalized Blogger modeling is the key of the personalized system to provide personalized services. The success of the whole system depends on the quality of the personalized Blogger model. On the other hand, separating the modeling and the application can make the system more expansible and robust. The application of personalized Blogger modeling is to establish a recommender system that can recommend correlative news and other Bloggers according to the Blogger' interests.