Research And Implementation Of Course Selection And Evaluation System Based On Collaborative Filtering And Decision Tree

With the deepening of education reform,many universities have developed a mobile educational management system to facilitate teaching management and improve teaching quality.However,in the face of a large variety of elective courses,the existing system rarely involves how students choose the right course according to their hobbies and learning needs,how to evaluate the teaching results reasonably and analyse the valuable information in the evaluation of teaching data.Therefore,researching and implementing reasonable elective course recommendation strategies,scientific teaching methods and correct analysis methods have practical significance in the educational administration system.Based on the existing research on elective course recommendation technology,this paper proposes a course selection method based on collaborative filtering of course weights,aiming at the problem of cold start and sparseness of course scoring data in the traditional course recommendation algorithm.The method fully considers the relevance of the course and the interest of the students in class selection,and effectively improves the accuracy of the course recommendation.Students' evaluation of teaching is realized based on the teaching evaluation index system,and a C4.5 decision tree classification algorithm is introduced to analyze the evaluation data.Finally,the paper designed and developed an elective course evaluation system based on iOS platform.The specific work of this article includes:1.Analyze the shortcomings of the course recommendation technology in the traditional course selection system,and propose a course recommendation method based on collaborative filtering of course weights for problems such as cold start and sparseness of the scoring matrix.Considering the correlation between the courses and combining the interests of students during class selection,build a student-course weight matrix,calculate the similarity between students,determine the neighbor relationship between students based on similarity,and give recommendations based on the weights of their courses result.The experimental results show that this method is obviously better than the recommendation technology in the traditional course selection system in the accuracy of the recommended courses.2.The C4.5 decision tree classification algorithm is used to analyze the student teaching data and calculate the attribute field of the maximum information gain rate in the data.Then branches are constructed according to different values of the attribute field,so as to construct the decision tree from top to bottom recursively..The rules are extracted from the decision tree and the relationship between the hidden attributes of teachers and the teaching quality is studied.The hidden information inside the evaluation teaching data is excavated so as to achieve the purpose of scientific and reasonable feedback teaching quality.3.Design and implement an elective iOS app based on Hybrid mode,using iOS Objective-C technology to implement mixed mode of selecting courses and Jsp/Servlet web development.Through the Maven tool,we call Mahout framework to collaboratively filter the parameters of the recommended algorithm.The jar packages in Mahout framework are excellent,they use the Recommender abstract interface to provided by Taste to implement the course selection algorithm proposed in the paper.