Rss-based Hierarchy Of User Interest Model Study

With the step of the 21st century, information needs are growing rapidly, the Internet has become the major media of the dissemination, exchange and sharing of information, its size is expanding with an alarming speed. A rapidly increasing number of global Web sites, and an increase in the number of websites have been expanded in scale. Various Web sites and complexity of information have increased rapidly. At present the website information and web database, which are the two main online information resources, remain in the information services through its network database online search and passive forms. The most obvious question of passive information services is that specific information often fails to reach the needs of specific users.In terms of the quality of the Web personalized serves, as the technical foundation of the Web personalized services, the technology of user modeling has a basic function. If there is no technology of user modeling that is mature enough, no matter how good the recommending technology is, e.g. information search and user clustering technology, it's impossible for user to enjoy mature enough Web personalized services.So far, all kinds of technology of user modeling remain not mature enough, there is also a lot work required to de done in the field of RSS-based technology of user modeling. For example, how to simplify the work quantity of the pretreatment the gradation model and then let the model capture the user interests more quickly; how to supply user with more personalized services; how to prevent the lost of the user interests in the negative update mechanism so that user model could capture the users interests more accurately and more fully.Therefor, this paper will try to establish a better RSS-based gradation user interest model, this will make a better foundation for Web personalized services.