Research And Implementation Of Exercise Recommendation System Based On Personalized Forgetting Model

With the development of Internet online education platform,knowledge tracing(KT)technology is more and more widely applied.By modeling the interaction of students’ answer history,KT can accurately predict the degree of students’ mastery of each knowledge point and further provide students with personalized learning plans.According to the existing research results,it can be found that the traditional knowledge tracking method doesn’t take the forgetting rate and learning ability of students into consideration when providing learning programs.In order to solve the above problems,this thesis constructs a personalized forgetting model for students from the perspectives of Ebbinghaus forgetting curve and Wilson confidence interval,and combines students’ knowledge mastery level and students’ historical answers to complete personalized exercise recommendations for students.The specific research contents are as follows:Firstly,this thesis designs a KT model based on deep learning.This model uses Ebbinghaus forgetting curve to construct students’ forgetting model,and provides exercise recommendations for students by combining with students’ historical answer interaction and knowledge mastery model.By comparison and analysis with the existing KT model,it is proved that the proposed model is superior to other KT models in terms of Accuracy(ACC)and Pearson correlation coefficient.Secondly,considering the different practice conditions of students in learning each knowledge point,Wilson confidence interval is introduced into the model of students’ personalized forgetting curve by combining the answer order of students’ practice questions,the number of learning questions and answer results.The cross-entropy logarithmic loss function was used as the objective function to optimize the parameters.The experimental results show that the prediction effect of the personalized forgetting curve model proposed in this thesis is better than that of the traditional forgetting curve.Finally,according to the above knowledge tracking model and the functional development needs of the system,this thesis uses Vue.js and sqlite and other mainstream technologies to build a set of student-oriented exercise recommendation system.The system realizes the functions of user information management,exercise,exercise management,record query and so on.At the same time,test cases are designed for different functional modules to complete the whole system performance test.