Research And Implementation Of Personalized Question Recommendation System Based On Knowledge Graph

In the era of fast-paced network information,modern living habits have undergone tremendous changes.In the past two years,the new coronal epidemic has continued to erupt at multiple points,which has brought many problems to offline education.Online education and learning have emerged due to its unique advantages.As a result,the number of questions on the question bank website is large,and the way of screening questions is simple and inaccurate.Students lack pertinence in choosing questions.The most important point is that they cannot effectively recommend the weak knowledge points,and ignore the knowledge closely related to the knowledge points.The problem is that the current knowledge cannot be mastered due to the insufficient mastery of relevant knowledge points,which ultimately affects the rationality and accuracy of students’ recommendation of questions.Based on the junior middle school mathematics knowledge graph,according to the process of students’ training and learning,this paper designs a personalized test recommendation method that combines the mastery level of students’ knowledge points with the correlation degree of knowledge points.Firstly,the junior middle school mathematics knowledge graph is constructed by combining manual and automatic methods,and then the knowledge graph is stored in the neo4 j graph database.The knowledge graph contains the properties of knowledge point entities and the relationship between entities,forming a hierarchical structure with precursor nodes and subsequent nodes.Secondly,the initial knowledge mastery level is obtained by DINA improved model algorithm,and the current knowledge mastery level is calculated by combining the knowledge migration rate generated by students in the learning process.Then the Trans R translation model is used to obtain the knowledge point vector and calculate the correlation between the two knowledge points entities combined with the importance of knowledge points.Finally,according to the current knowledge level and the correlation between knowledge points,the recommendation strategy is determined to recommend to students questions that meet their needs.The experimental results show that this method makes the results recommended by the test more accurate and reasonable,can accurately capture the blind area of students’ knowledge,and more effectively improve students’ learning efficiency.In this paper,the personalized test question recommendation method based on knowledge graph is used to realize the personalized test question recommendation system based on knowledge graph.The development of the system uses SSH framework,and the database is built using My Sql and neo4 j graph databases.The system mainly includes user center,online learning,recommended questions and knowledge point retrieval.The personalized question recommendation system effectively provides suitable questions for students’ weak knowledge points,thus improving students’ knowledge level.