Research And Implementation Of Exercise Recommendation System Based On Multiple Factors Knowledge Tracing

The rapid development of internet technology broadens the development mode of intelligent education platform.Classroom education enables students to swim the ocean of knowledge without the limitation of time and place.In addition,students can also mine necessary information from massive historical learning data,which provides convenience for personalized teaching of students and teachers and improves the learning efficiency of students.The previous recommendation algorithm has made a lot of attempts to consolidate learning results by recommending exercise for students,so that students can better develop the next learning scheme and plan,and improve their learning efficiency.Although the existing exercise recommendation algorithm can recommend exercise for students to some extent,it still has certain limitations: collaborative filtering algorithm can recommend exercise by looking for similar users and exercise,but ignores the mastery of knowledge of students,so it is difficult to recommend appropriate exercise to students timely according to their mastery of knowledge.However,the method of knowledge modeling starts from the individual level of students and ignores the characteristics of similar students in most cases.To solve the above problems,this thesis proposes a personalized exercise recommendation method combining multiple factors knowledge tracing model and collaborative filtering algorithm,which takes into account both the knowledge level of individual students and the knowledge level of similar students.The main ideas are as follows:(1)The multiple factors knowledge tracing(MFKT)combines multiple factors that affect the mastery of knowledge of students: The knowledge association of exercise and the knowledge difficulty of exercise updates the proficiency of knowledge for students,using dynamic key-value to update the mastery of each knowledge concept for students through memory network,then use long and short term memory network to get the knowledge level matrix after forgetting processing,so as to model the knowledge level of students.(2)Combined with the collaborative filtering algorithm,the knowledge level matrix of similar students is obtained through the knowledge level vector of students,and the difficulty range of exercise is determined according to the degree of students’ mastery of knowledge,so as to recommend the exercise of the corresponding difficulty range for students.(3)Finally,combined with the demand analysis of the system,design and implement the multiple factors knowledge tracing personalized exercise recommendation system research and implementation.The system is divided into the student end and the teacher end.The student end has the function of test,error record,knowledge visualization and exercise recommendation,and the teacher end has the function of subject management and test paper management.Considering the high efficiency of the system operation,the front-end and back-end are designed respectively.Vue framework is used to display the front-end function,and Spring Boot+My Batis framework is used to design the backend function.