Research And Design Of The Learning Platform Based On Personalized Resources Recommended

With the rapid development of the Internet, the Internet provides people with abundant learning resources, and has become the main way for people to obtain information, is also one of the most common ways. With the increase of Internet Information, the information search technology is facing great challenges. The traditional search technology already can’t adapt to or meet the personalized requirements of people. It’s more and more difficult to reach what you want and interested network resources from the Internet. The recommendation technology was born, has developed rapidly and showing a strong vitality.Among the numerous Recommendation Technologies, collaborative filtering recommendation technology is one of the most widely used. Collaborative filtering recommendation algorithm can help the users find potential interests and hobbies, give recommendations, let users find resources that suited and interested quickly from the plenty of Internet resources to meet the personalized needs of users. However, by reason of the collaborative filtering algorithm needing to rely on users’ ratings of resources to calculate the similarity, the disadvantages of collaborative filtering algorithms have been evident, such as cold start and sparse data problem.In the field of education, there are a lot of network teaching platforms that contain abundant learning resources, but they don’t pay attention to the learners’ individual needs and present the same content to the learners. In this paper, I apply the personalized recommendation technology into the field of education to help learners find the resources they are interested in diverse learning resources, and push for them to learn.The main work of this paper is divided into the following parts: through the comparison of common personalized recommendation system, the most widely used collaborative filtering recommendation algorithm is choosen; In order to make up the deficiency of collaborative filtering algorithm, the improvement scheme is proposed: the mixture algorithm made up by content-based recommendation and collaborative filtering recommendation algorithm can improve the quality compared with the traditional collaborative filtering algorithm; using social tagging systems annotate the learning resources, setting up the learning resources model, and then calculating the similarity between the resources make up the problem of data sparseness effectively; Collecting user data, for basic information of the learners’ implicit data, setting different grade for different learning behavior of learners, indicating different preferences set up learner model; After the detailed design of the various learning platform modules, the implementation of the platform system was gone into detail, and the front desk and backstage module was also tested.