Based On Students Of Campus Card Consumption And Data Mining Application Research

With the deepening of the information construction of colleges and universities,the university in the use of information gathered and accumulated data,have a higher demand,"campus one-card",a Shared database,and so on the construction of the campus information platform for global data mining provides a data basis.Based on the in-depth analysis and study on the basis of data mining technology,applied to the "campus onecard " water consumption data analysis,dug deeper understanding of students' consumption level and consumption pattern,the auxiliary funding for poor management decision analysis has a certain practical significance.This paper studies data mining related technologies,and focuses on the analysis of k-means method,Apriori association rules mining method,and related applications in campus card consuming water mining.First,in view of the campus one large amount of water consumption data and dimension of more features,select the data source,data preprocessing,focuses on how to make the campus card water table and the related data in the table information data through data cleaning,data integration,and variable selection and transformation,it is ready to become available for subsequent data mining process operation on the basis of the data.Second,using literature retrieval method to find the consumption level of the index,based on the characteristics of these indicators and student card consumption students consumption level index model is set up,with the three indicators-student canteen consumption intensity and student canteen meal rate,comparing students' consumption level difference,the k-means algorithm are used to cluster all students consumption level,students' consumption level can be divided into four major categories.Finally,using the k-means clustering results are Apriori association rules mining frequent rules,according to the result of mining association rules in the knowledge that is used to solve the problem of poor students evaluation work practice,for the poor students recognition mechanism are also proposed.Studies have shown that applying data mining to campus one of data analysis,and assistant decision-making to school help poor students recognized the results of analysis is feasible,and the knowledge of the mining results for early warning,monitoring deemed poor students lay a solid foundation.