The Design And Implementation Of Campus Card System Data Analysis For Universities

The application scope of card systems in colleges and universities is becoming wider and wider.Cards everywhere generate data.The amount of data generated every day can reach more than 100,000 transactions.Hence,the amount of data accumulated over time is even greater.The data describes students' daily lives and consumption habits.There is no doubt that these data contain valuable information.The analysis and mining of campus data is of great significance for timely checking learning situation,guiding effectively students' study life,and improving university management.This thesis uses data provided by S colleges and universities,based on the card system of colleges and universities,and the actual business requirements and card related business,in-depth researches on data multi-dimensional analysis technology,data visualization technology and data mining technology,and finally completed the design and implementation of campus card system data analysis for universities.The work has carried out preliminary exploration and analysis for the data mining of the university card system,and also prepared a basis for the establishment of a decision support system in the future.The contributions of this thesis are as follows:(1)Based on the professional internship experience of the card management platform and card manufacturing center and the need for the construction of a smart campus,design an analysis framework,propose the theme of the analysis,and design a multi-dimensional analysis model for consumer behavior and a multi-dimensional data model for library access.And based on SQL Server to implement multi-dimensional analysis,including data ETL,design access control and multi-dimensional slice rotation.Several behavioral patterns are found through visual observation.(2)The clustering data mining model is designed and implemented.Through clustering analysis it was discoveried that students can be divided into five categories according to their consumption habits: General,High Consumption,"Stay at Home",Successful Learnner,and Frailty.Students who do not eat or bath regularly go always to the infirmary,and the health situation of postgraduate group is weak.(3)Designed an association rule data mining model,and used the Microsoft association rule algorithm to implement restaurant window binding recommendations and window merge suggestions;designed a student daily activity behavior sequence clustering mining model,and implemented student clustering analysis algorithm to realize student behavior sequence mining.It was found that regular dietary groups generally do not go to the infirmary and other student behavioral sequence patterns.