Visual Analysis Of Campus Card Application Data

With the introduction of the concept of smart campus,the mining of educational data has become a hot issue for people to study.Campus card application data,as the business data that best reflects the life of students'campuses,plays a vital role in the construction of smart campuses.Aiming at three research problems of campus card application data storage and management,data mining and analysis,data visualization,this paper designs and implements an analysis and visualization system based on campus card application data to help teachers and students understand the implicit data.Regularity provides decision support for students' campus life and management of management personnel.Firstly,this paper designs a topic-oriented data warehouse for the characteristics of campus card application data,and aggregates multi-source heterogeneous data at different levels for easy query and analysis.Secondly,this paper analyzes the campus card application data from the perspectives of data mining and traditional statistical analysis.From the perspective of data mining,the automated ARIMA model is implemented and applied to the forecast of human traffic in campus hotspots.An improved k-means algorithm is proposed and applied to the clustering analysis of consumption features.A campus card application is proposed.The student achievement prediction model of the data extracts 7 kinds of characteristic indicators that can affect the students'performance.The random forest algorithm is used to predict whether the students have the risk of hanging the subject,and the validity of the results is proved by experiments.From the perspective of traditional statistical analysis,statistical analysis is carried out on consumption distribution characteristics,student behavior similarity,term bills,hot merchants and consumption details.Finally,specific visualization models and interactions are designed and implemented for different analytical topics.