Research On Personalized Learning Based On Collaborative Filtering Technology In E - Learning System

With the rapid development and popularization of the Internet, more and more people prefer to use the modern E-learning system to learn. But in the network there are too many learning resources, users are confused to these huge learning resources, and they can not accurately and efficiently find the resources which they really need. In order to help users to find the resources they need as quickly as possible, personalized learning recommendation technology has come into being. E-learning systems and recommend technology has good prospects for the development and application, and it has become an important research content, including personalized recommendation system is designed to meet the specific needs of different users’ requirement, which also is a branch of research.Firstly, this article gives a general overview of a personalized learning system, and use application ontology to build the knowledge systems, according to the system’s requirements of the learning resources environment. The use of ontology can promote to build knowledge system and organize the learning resources. So this E-1earning system provides a good environment for personalized learning. Also teachers can freely use the application ontology to design the structure which is needed in teaching knowledge, and have the freedom to choose learning resources and application ontology what are linked to prepare for the completion of the learning content. Then it introduces conceptual level of the learning resource that is based on application ontology, which paves the way for making personalized learning recommendations. The recommendation is based on collaborative filtering recommendation algorithm in order to improve the quality of recommendation. Due to the lack of traditional collaborative filtering recommendation algorithm, this article gives an improved collaborative filtering recommendation algorithm, which compares with the traditional collaborative filtering recommendation algorithm, its main advantages are followed by such aspects: first, using a tool named concept of hierarchical tree to easily find the hidden relationship between resources. Relying on this analysis of species level tree, the users’ scores of resources become closely connected because of the implicit relationships, which can help to find a more accurate recommendation system neighbors, then recommender systems can produce more accurate recommendation, effectively improve recommended targeted. Second, when the improved algorithm is in the calculation of the target users’ neighbors, it adds the time element which gives weight target users visited recently in the project, reflecting the changing trend about needs of users. Neighbors and resources which are found by this way are more similar with target users’ demands. Then in such a relaxed personalized learning environment, user autonomy to determine their own needs and their own way of learning content, and the system records the performance of users and the extracted effective learning strategies for other users by reference. This approach could theoretically infinite reserves expanded learning strategies to adapt to the needs of different users. It also provides accessibility features for teachers to design reference learning strategies.