Research On Collaborative Filtering Algorithm Orientating Personalized Recommendation Of Educational Resources

With the rapid development of information technology,people increasingly pay attention to educational informatization.In the process of educational informatization,the imperious demands of learners is realization of personalized learning.Personalized recommendation technology can help personalized learning,it can improve the learner's learning efficiency.Firstly,the Article elaborates the background and current states of personalized learning and introduces related theory and key technology.Secondly,the Article researches personalized recommendation models of educational resources,which contains user model,educational resources model and knowledge points model.Meanwhile,it elaborates the basic thought of collaborative filtering algorithm,and discusses method of similarity measures in collaborative filtering algorithm.The article introduces two process modes about the algorithm,which are based user and item.Thirdly,the article focuses on the matrix sparsity problem in collaborative filtering algorithm,such as when the algorithm searches similar learners,it maybe can't find neighbor and causes the inaccurate result.So in the article,it designs algorithm of collaborative filter base on SVD.The algorithm search the neighbor after reducing matrix.Fourthly,the Article elaborates the supporting platform named mobile independent school and use the educational data to do some experiments.The experimental result indicate that the collaborative filtering algorithm can provide accurate prediction.The improved algorithm is better than original algorithm.