Research On The Method Of Student Profile Construction And Performance Prediction Driven By Data

With the popularization of information technology in education,more and more attention has been paid to Educational data mining education.Data on students at school is also accumulating.Mining and analyzing these data can help schools better understand every aspect of every student.To provide students with better learning experience and campus services.Therefore,this article constructs the student profile and the performance prediction model,enhances the student informationization and the personalized service level.The main research contents of this paper are as follows:(1)Student behavior analysis based on improved fuzzy clustering.In this paper,several methods are used to improve the fuzzy c-means clustering algorithm.First,the initial clustering center is determined by Gauss density function.Then replace the Euclidean distance with the density-sensitive distance.The contour coefficient and elbow method are used to determine the optimal cluster number.Finally,the effectiveness of the improved clustering algorithm is verified.Using the improved clustering algorithm to analyze the three dimensions of students,including consumer behavior clustering,learning behavior clustering,life behavior clustering.The behavioral characteristics of different types of students were summarized.(2)Performance prediction method based on multi-attention mechanism.The neural network model is used to combine the students’ behavior characteristics and historical performance in the past year.A parallel set of attention mechanisms was used to calculate the weights of behavioral traits on the last two semesters’ grades.Another attention mechanism was used to extract the relationship between performance and behavior over two semesters.Finally,based on the characteristics of the integration of the next semester grades for more accurate prediction.The experimental results show that the model improves the accuracy of performance prediction.(3)The design and implementation of the student integrated portrait system.Using clustering results and grade prediction model to construct the label of student portrait.Design and realize the functions of student behavior management,student comprehensive profile and grade prediction.Finally,Echarts is used to realize the large-screen visualization of student profile.