Research Of Personalized Exercises Recommendation Based On Knowledge Points

As the most important tools of solving information overload problem,recommender systems gradually change people's way of lives.Personalized exercise recommendation is a branch of recommendation system in the field of education.In recent years,it has attracted a lot of researchers to carry out a wide range of researches.In this thesis,personalized exercise recommendation methods based on knowledge are proposed,and the main issues are as follows:(1)To deal with the problems that the cognitive diagnosis model can not grasp the probability of knowledge points,we put forward a probability model of student knowledge.Then TopN personalized exercise recommendation algorithm is proposed.Experiment results show that the algorithm has a accuracy.Comparing to the cognitive diagnosis model,the accuracy rate increase by 10.1%in the FrcSub dataset,2.84%in the Mathl dataset,In the Math2 dataset,not only does the accuracy increased by 1.39%,but also the operating efficiency is 12.1 times faster then the cognitive diagnosis model.(2)We propose a weight graph that characterizes the hierarchical relationship of knowledge points,first we construct a hierarchical relationships between knowledge points and the loss rate matrix of students' knowledge points,and then we solve the problem of low recommendation accuracy caused by the sparseness of data between students'knowledge points.Experiment results show that the algorithm has a good effect on the self-built data set.Compared with personalized recommender systems of knowledge points without hierarchical graph,the recommendation algorithm lead an increase of 12.35%in accuracy,5.49%in recall rate and 11.91%in F1 level.(3)In order to solve the problem that the existing evaluation criteria can not accurately reflect the students' grasp of the knowledge points,a new evaluation criteria based on knowledge points is proposed in order to recommend more accurately of the students'weak knowledge points.In order to verify the rationality of the proposed method,we design and implement a prototype of the recommendation system based on knowledge points.The experiment results show that this system can be used as an important auxiliary teaching method,not only to make up for the different levels of students' knowledge loopholes,but also to improve the efficiency of students' autonomous learning.