Research On Recommendations Of Personalized Learning Resources Based On Knowledge Graph

With the rapid development of network technology,online learning has been accelerated in the field of education and learning through the Internet becomes a common phenomenon.Therefore,it puts f orward higher requirement for the depth and breadth of the learner's knowledge.Although the developme nt of artificial intelligence technology has improved the degree of resource acquisition for learners,how t o better help the learners is still an urgent problem to be solved in the context of massive learning resourc es.It includes how to make learners clear their learning direction in a limited time and recommend appro priate learning resources to learners,which aims to help them improve their learning efficiency.The key to overcoming the “losing the direction of learning” and the “cognitive overload” in network learning is how to help learners find suitable learning path among the massively growing learning resources so as to provide learners with accurate and personalized test recommendations.In view of the above problems,this paper studies the learning path recommendation and personalized test recommendation in learning resources.The following is the main work of this paper:(1)The related theories of knowledge graph are expounded,and the relationship between knowledge points and their properties and the construction method of knowledge graph are studied.Construct knowledge graph based on the knowledge points in the curriculum and study how to construct knowledge graph based on knowledge and dependencies among knowledge points.(2)According to the dependency relationship between knowledge points,the individual contribution value of the knowledge point is defined.Then proposed a centrality calculation method based on Knowledge point contribution value and Knowledge Point Value and Contribution Value respectively.The validity of the algorithm proposed in this paper is verified by comparing the simulation experiment with the existing knowledge point centrality calculation method.Finally,the knowledge points of different knowledge levels in the knowledge graph of knowledge are sorted,and the descending order of centrality is chosen to recommend learning paths for learners.(3)In order to solve the problem of blind learning and low efficiency of knowledge overload,a personalized test recommendation algorithm based on knowledge point graph is proposed.First,according to the relationship between questions and knowledge points,establish a knowledge matrix of questions and through the learner's test answer record,calculate the score of knowledge points.Calculate the new knowledge points loss rate by value of contribution between knowledge points and old knowledge points loss rate,which compare the set threshold in order to select learners to master weak knowledge and recommend a test for the learner.Experiments show that the test recommendation algorithm proposed in this paper can improve the accuracy and effectiveness.