EduExplorer: Visual Analysis System Based On Campus Behavioral Data

The understanding the relationship between students' behavioral patterns and their academic performance has always been a key in personalizing Education.A large amount of papers have previously demonstrated that students' physical characteristics(such as height and weight,etc.)and psychological characteristics(such as personality and IQ)are related to students' academic performance.However,the vast majority of psychological and pedagogic research results are based on subjective questionnaire surveys and psycho-metric scales.In addition,most of the experimental samples are very limited.Therefore,there may be some deviations in the personality indicators obtained.With the advance-ment of education informatization,educational data are rapidly emerging.Although anal-ysis of electronic data can solve this problem to some distance,however the diversity and accuracy of educational data is an increasingly challenges.There are a lot of research using the data mining algorithm to obtain education knowl-edge in the data.Although the data mining algorithm can achieve a more objective conclu-sion,analyst can hardly understand the data from a global perspective.As a result,some researchers are utilizing visual analysis to help users to analyze educational data in visual way,such as PeakVis and VisMooc,a visual analysis system designed to help MOOC instructors to have a better understanding of their students'behaviors.Visual analysis can enable analysts to understand data from a global perspective.As a result,visual analysis is also playing an increasingly important role in various fields.The main research contents are as follows:(1)Analyzing a large number of students' card data,use real entropy to quantify stu'dents' order and verify the reliability of experimental data.By analyzing the electronic data of students' daily behaviors,subjectiveness of the traditional psychology question-naires in the psychological scale were avoided.In addition,the data analysis of 18,960 students also avoided the problem of small sample in traditional psychological research.In order to analyze the student's personality,this article introduced the theoretical founda-tion of the subdivision traits' achievement and order in the conscientiousness of the five personality theories in personality psychology,and use the real entropy to describe the order and cumulative number to characterize the achievement.So far,this article is the first to use student card swiping data as a data source to analyze student behavior.(2)Design the visual mapping algorithm of the system,the layout algorithm of the view and the interaction algorithm between different views,and on this basis,the visual analysis system is implemented.This paper designs visual mapping algorithm for swiping data,realizes the mapping and transformation process from data to visual elements,de-signs layout algorithms for each view to present data,and designs interaction algorithms between each view.Finally,on this basis,the visual analysis system eduExplorer is im-plemented.This system is used to explore the temporal and spatial characteristics of the student's credit card data,and the availability of the system is proved through the analysis of several cases.