Research On Learning Analysis And Personalized Exercise Recommendation Method Based On Neural Cognitive Diagnosis

In recent years,online education systems such as Massive Open Online Courses(MOOCs)have become increasingly popular,which provides the possibility of personalized learning for students.Personalized learning requires learning analysis to pay attention to the individual differences of students,and to match accurate learning resources for different students based on students’ knowledge level.Cognitive diagnostic assessment is designed to measure the student’s knowledge level.However,the existing cognitive diagnosis models usually use artificially designed functions to mine the students’ practice process.These functions are usually relatively simple and cannot well capture the complex relationship between students and exercises,which affects the diagnosis effect.In addition,in order to better realize personalized learning,after accurately diagnosing the knowledge level of each student,it is necessary to provide students with personalized learning remedial services,such as personalized exercise recommendation.As an important learning resource,exercises play an important role in the learning process of students.However,for students,it is difficult for them to choose exercises of suitable difficulty due to the differences in their own knowledge levels and the huge resource of exercises they face.Therefore,this topic selection is based on the product upgrade of an education platform as the research background.This article will conduct in-depth research and exploration around the study of neural cognitive diagnosis-based learning analysis and personalized exercise recommendation methods.The main contributions are as follows.1.Based on the Neural Cognitive Diagnosis(Neural CD)framework,an Enhanced Neural Cognitive Diagnosis(ENeural CD)model is proposed,which improves the fit of the complex relationship between students and exercises,improving the diagnosis effect.2.The cognitive diagnosis results obtained by the ENeural CD model are effectively applied to predictt students’ score.And a personalized exercise recommendation method combining neural cognitive diagnosis and neural collaborative filtering(NCD-NCF)is proposed.This method first builds a Neural Matrix Factorization model incorporating Personality Characteristics of students’ knowledge level(Neu MF-PC),which is used to predict the probability of students answering the exercises correctly and then recommends reasonable exercises for each student by setting the difficulty threshold.3.In view of the above two research works,a large number of experiments have been carried out on real datasets.The experimental results prove the effectiveness of the ENeural CD model and NCD-NCF personalized exercise recommendation method proposed in this paper.