Development Of Smart Education Platform Based On Heterogeneous Multi-source Learning Data

With the rapid growth of the Internet,more and more online education platforms have grown.The use of online education platforms has greatly facilitated the study of students and the management of teachers,and has revolutionized educational concepts and educational thinking,but how to use them in the teaching platform.Doing a good job of personalized recommendation and improving students' learning efficiency is still a problem worth studying.Faced with the variety and existence of data in the education platform,this thesis proposes the organization and processing of heterogeneous multi-source data based on the efficient organization and processing technology based on big data.The data fusion technology and heterogeneous data verification technology are constructed,and a dynamic learning model of heterogeneous multi-source is established.By analyzing the correlations in the subject Knowledge Graph,this thesis proposes the method of knowledge point metrics and node metrics,and obtains the knowledge point importance ranking algorithm,which lays a foundation for the research of recommendation system based on Knowledge Graph.In view of the low efficiency of the traditional recommendation algorithm in the education platform,the algorithm is difficult to implement,the algorithm is difficult to implement,the recommendation is inaccurate and the problem features are not easy to extract.This thesis proposes a knowledge recommendation algorithm based on Knowledge Graph,according to the content of the course,The Knowledge Graphs of different disciplines are obtained,and the knowledge point importance sorting algorithm proposed in this thesis is combined to calculate the order sorting sequence of knowledge points importance.New users in the education platform can directly use the knowledge point importance sorting algorithm to get the recommended list for problemrecommendation.For users with existing data records,according to the constructed user behavior record library,the correct problem sequence based on the students themselves can be obtained.The problem sequence and the knowledge point-question distribution matrix,and calculate the knowledge point score order rate sequence,combined with the recommended interval function and the knowledge point importance sorting sequence,construct a training set and get the recommended list of exercises recommended to each user.This thesis implements a number of smart education service platforms,and implements the function of the knowledge recommendation algorithm based on Knowledge Graph proposed in this thesis.It improves the intelligent push in the smart education service platform and provides students with personalized problem recommendation.Services to improve the efficiency of student learning.