Decentralized Online Learning Based On AI Systematic Research

With the rapid development of artificial intelligence,online learning has developed rapidly in the field of education too,and the Internet has been widely used in learning.Although the development of artificial intelligence technology makes it more convenient for users to obtain learning resources,in the context of perplexing mixed learning resources,how to let users clarify the learning direction within a limited time and intelligently recommend resources to users to help users improve learning efficiency and change To have a more comprehensive grasp of relevant knowledge is an urgent problem that needs to be solved.Because there are too many learning resources in online learning,there are problems that can not find proper learning methods and suitable problems.The main problem to deal with these problems is how to make the online learning system find a suitable learning path among a large number of growing learning resources to provide users with precision Recommended for personalized resources.In response to the above problems,this article studies the different characteristics of each user,and implements personalized and resource recommendations for each user's different characteristics.The following is the main work of the paper:(1)For the security of learning resources in online learning,distributed storage is implemented for various data resources,and the problem of how data sets are divided in distributed databases is discussed.Finally,the segmentation of lexical semantic similarity and node density is used method.(2)Aiming at different characteristics of different users,a ranking learning recommendation algorithm based on knowledge graph is proposed.The algorithm uses ranking learning to build a feedback feature model,taking into account the user interest migration problem,and establish a user migration model fusion,and then fusion with the basic feature model to build a hybrid model,through the final hybrid model to achieve Top-N recommendation.To improve personalized recommendations.(3)To test the content learned by users,a personalized test recommendation algorithm based on knowledge graph is proposed.Firstly,a method for constructing knowledge graph of knowledge points is proposed,and then a heuristic personalized test question recommendation method based on historical test score data is proposed.Finally,the testshows the effectiveness of the algorithm.