Learning Path Recommendation Based On Knowledge Graph

With the rapid development of the mobile Internet,online learning has also developed in the field of education,and learners have begun to switch from traditional learning methods to online learning.The Internet has everything available for learning resources.Learning resources in any subject and field can be obtained.The massive learning resources on the Internet bring problems such as "cognitive overload" and "knowledge lost" to learners.Although the development of artificial intelligence technology has improved the access to learning resources,it is quite difficult for learners to master knowledge points in a short period of time in the context of online learning resources.In recent years,recommendation algorithms have achieved good results in the education field.They can recommend content of interest to learners and filter redundant resources.However,learners still face the confusion of excessive recommended content that cannot be sorted out and knowledge fragmentation.How to help learners improve their learning efficiency and obtain more comprehensive professional knowledge is a problem that needs to be solved today.In response to the above problems,this paper uses the knowledge graph combined with neural network to study the learning path recommendation.For the problems that network learning resources make learners' knowledge lost and low efficiency,a model of learning path recommendation based on knowledge graphs is proposed.First,construct a knowledge graph based on the obtained entities and relationships,use graph database for storage and visualization,and then use graph neural network and gated neural network to mine the implicit relationship between users and knowledge points,and obtain the relationship between each knowledge point and the graph.The degree of importance,and finally the knowledge points are evaluated,and the learning path of the online learning resources is obtained.Through comparative experiments,the validity of the model proposed in this paper is verified.This article uses the method of user profile to dig out learners' mastery of knowledge points.A recommendation algorithm based on tag weight is proposed.First,the learner is profiled with the user portrait,and then TF-IDF is used to calculate the comment information,and finally the label weight matrix is formed to recommend learning resources or learning knowledge points for the learners.For network learning resources,through the neural network model proposed in this article,the knowledge graph construction method is used to evaluate the model proposed in this article through two evaluation methods.After experimental analysis,the learning path generated by the model in this article is better than the learning path generated by other algorithms..The use of user portraits and tags is used to portrait learners,which improves the learning path recommendation for learners.