Research And Practice Of Adaptive Learning Path Generation Technology Based On Knowledge Graph

With the rapid development of the Internet,online education has also entered the public eye.Education is related to the national economy and the people's livelihood.Adaptive personalized learning as a more efficient and better allocation of social resources is becoming the direction of future education.This model models participants in education,teaching materials,and data generated during the interaction process,analyzes the learner's personalized characteristics,and provides adaptive learning content and solutions based on the changing characteristics.The adaptive learning system faces the following major problems:firstly,how to represent various entities in the online education field,secondly how to complete the missing information(ie,the relationship between the entities in the education field),and finally how to implement personalized adaptive recommendation.Aiming at the above three issues,this paper will carry out related research and work in the following aspects:First,there are many types of objects in the online education field,and the relationships are complex.In this paper,a scheme for modeling all objects in the field of education as a knowledge graph is proposed.With the powerful knowledge representation and reasoning capabilities of the knowledge graph,the data in the education field can be fully mined.Secondly,the relationships between entities in the online education field cannot be all established manually,so an automatic labeling mechanism needs to be designed to complete the knowledge graph.In this paper,a multi-label classification algorithm based on attention mechanism is proposed to solve the problem of imperfect relationship between topics and knowledge points.The labeling of the problems is regarded as the classification task to predict the knowledge points related to the topic.Finally,most online learning platforms provide the same learning solution for all students.In this paper,an adaptive learning path recommendation algorithm based on network embedding and historical learning effects is designed to recommend a personalized learning path for each student.In this paper,an adaptive learning system is designed and developed.In addition to the basic online learning and testing functions,the system can also visualize the relationship between knowledge points in different disciplines more intuitively.At the same time,a multi-label problem-based algorithm based on attention mechanism is used to integrate and label related resources in the system.An adaptive learning path generation algorithm is used to recommend appropriate learning paths for students,making the system more intelligent.The relevant tests on the adaptive learning system have all passed,which proves that the system developed in this paper is usable and effective.