Design And Implementation Of Mobile Educational Administration System Based On Collaborative Filtering Recommendation Algorithm

Due to the rapid development of Mobile Internet Technology,the traditional educational administration system in universities cannot meet the daily learning needs of students,and most of the applications of educational administration system need to be connected to the mobile terminal gradually.With the deepening of China’s education reform,colleges and universities will organize a variety of elective courses for students to study in order to make students have a rich learning life.When students are faced with a large amount of information about elective courses during the course selection,they usually need to spend a long time to select the courses they like.Aiming at these problems,this paper designs and develops a mobile educational administration system based on collaborative filtering recommendation algorithm,which provides students with teaching administration information on the mobile terminal and recommends elective courses that meet students’ own needs and characteristics during course selection.This paper studies the key problems of NGCF(Neural Graph Collaborative Filtering)recommendation algorithm in application scenarios firstly,such as low recommendation accuracy and cold start,and then develops "Mobile Educational Administration System Based on Collaborative Filtering Recommendation Algorithm" by combining advanced system development technology.The specific research work and results are as follows:(1)This paper proposes a set of students’ course selection data sets,compare and use several graph convolution recommendation algorithms on this data set,judge the feasibility of NGCF collaborative filtering recommendation algorithm as the system core recommendation algorithm,and then use NGCF algorithm to generate initial recommendation model.In view of the shortcomings of the NGCF algorithm in this application scenario,a method of adding student credits and the college fusion layer is proposed,the candidate set is filtered according to the students the current credit situation and the current course history of the college to generate a more accurate recommend model;A method of solving the algorithm cold start at the system application level is also put forward.(2)Combining the improved collaborative filtering recommendation algorithm,the requirements of the system are firstly analyzed,and the overall architecture,security,functions and database of the system are designed.Then the development technology is used to implement the system functions,and the system is combined with WeChat public platform to realize the access of mobile terminal.Finally,this paper also carries out unit test and performance test on the system,and finds out the system’s errors timely in the development process,so as to facilitate the system’s later maintenance.