Application Research Of Data Association Analysis And Mining Technology In Student Information

With the development of the country,society,science and technology,the demand for talents in the country and society has become more enormous.Colleges and universities are the cultivators of talents.Therefore,colleges and universities need scientific management to improve the quality of teaching and guarantee to delivery of national and social talents.Based on the school programs and students,this paper from the perspective of school programs arrangement,individualization of students' personality,and factors affecting students' performance,uses data mining related technologies to explore how to improve the quality of teaching in schools.For the school programs setting,this article mainly studies the relevance of the programs;for the individualized training of students,this paper takes the student achievement as the research object,and proposes the student group division method based on the improved KMeans algorithm.This paper uses decision tree algorithm obtains the key factors affecting students' performance software system was developed to implement the various algorithms mentioned including:(1)A curriculum relevance analysis algorithm based on Euclidean distance is proposed.The algorithm calculates the relevance of the course by measuring the difference of the course in the different dimension from the perspective of distance;the relevance of the course based on cosine similarity to analysis algorithm,by measuring the cosine value of different vectors,which calculates the relevance of the course from the perspective of the vector;the relevance analysis algorithm based on the correlation coefficient from the perspective of centralization evaluated relevance of different courses.After pre-processing the course results,two different data sets were selected to experiment on the three algorithms.The results show that all three algorithms can calculate the relevance of the course and provide scientific data support for the school curriculum.So that the school can scientifically verify and optimize the curriculum arrangement,three algorithms are compared from calculation results and theory,among which the correlation coefficient correlation analysis algorithm works best.(2)A student group partitioning method based on improved KMeans algorithm is proposed.The standard KMeans algorithm is improved on the initial clustering,which can improve the initial clustering selection in the standard KMeans algorithm.Disadvantages effects.And selected four data sets based on the number of different students and courses to conduct experiments,and the results were analyzed to find out the characteristics of students,which can provide a reference for the school to achieve personalized training.To enable teachers to train according to the characteristics of students(3)The students' performance factor analyze method based on decision tree algorithm is proposed.The questionnaire is used to collect student information.The decision tree algorithm is used to establish the corresponding decision tree based on the information gain.The key factors affecting students' performance are obtained.Provide valuable reference information for improving student achievement and student counseling.(4)Developed a Java-based student information analysis system,including course association analysis,student group division,and student achievement analysis module,to implement the algorithm used in this paper.