Research Of The Update Services On Regional Education Informationization Resources Repository

With the Internet technology's becoming more and more advanced, popular and perfect, gradually people s'life, work and study cannot be limited by time, space and region. The network has become indispensable parts of people's life and study. In the field of education, the speed of education informatization has been irresistible. Meanwhile, the boosting of education informatization process has put forward higher requirements to informatization education platform and education resources repository.Currently, there are many kinds of websites about education, these network resources give people rich knowledge and great convenience, but it exposes some problems. The resources which are in "explosive" growth sometimes do not meet users'real needs. In order to solve the problem, "personalized service" is studied and used by more and more people. Personalized service provides resources to users by analyzing user's Personality and preferences of which users might be interested in. The author produces an idea when learning from personalized service technology:for users personalized service, it can only provides services passively. If you don't solve the problem from the resources repository itself, it is impossible to improve the resource repository. Eventually the one who can solve the problem from technical aspect is nobody but the resource manager himself. This paper starts with solving this problem, and provides a kind of personalized service for resource managers.Focusing on the resources update service, this paper introduces the principles and executive process of resources update services model by collecting users'information, the establishment of user interest model, the establishment of resource model, content and collaborative filtering, model calculation etc. During the implementation, for the user's interest model and resource model, the author uses the expression method based on vector space, forgetfulness algorithm based on Fibonacci; for collaborative filtering, the author adopts clustering algorithm in data mining. After the model is built, through calculating Euclidean distance between models, we can push the corresponding information for managers. This method fundamentally solves the problem that resources cannot satisfy user's need.Finally, by simulation experiments, this paper verifies the feasibility of repository update service system, provides strong support and assurance for the update service's further development and research.