Research And Implementation Of Campus Card Application Data Analysis System

With the advent of China's comprehensive information age,more and more items have been applied to the "one-card system" of colleges and universities,including restaurants,bathrooms,library access control,and book lending.There is a huge amount of data stored in the database of the one-card system,but there are a lot of important information that can reflect students' behaviors and habits.With the advent of the big data era and the popularization of digital campus,more and more scholars began to pay attention to this part of data in universities.This paper is to study and complete an application based on campus id data analysis system,the main research and using clustering algorithm and association rules algorithm,on the basis of southwest university of science and technology IC card data,data analysis and processing.On the basis of existing K-means algorithm for K value as well as the selection of clustering center to make improvement and define it as VW-K-means algorithm,the improvement in algorithm first initialization phase to join the Voronoi diagram,then the data is simple processing,eventually make its own to select clustering center and K value;In order to ensure that in the process of operation can then clustering of data objects quickly to the nearest cluster,but also more accurate in order to be able to distinguish different between each data to like,in the improved algorithm based on Voronoi diagram and its criterion function is improved using the weighted average method,get it.Compared with the K-means algorithm and TDKM algorithm,the clustering accuracy is evaluated.After the algorithm is improved,the paper puts forward the system detailed design requirements,and completed the design of the data analysis system,including data layer,data processing,data warehouse and data service layer.Using SQL Server 2005 Analysis Services for data source integration,transformation,loading,and then build the data warehouse,the IC card consumption data for Analysis,finally uses data visualization technology,the Analysis of the data results visually displayed in front of the user.In this paper,the system functions mainly include bathroom analysis,restaurant consumption,library access and borrowing analysis,and poor students' prediction of four modules.This paper combines data warehouse,OLAP(Online Analytical Processing),data mining and other technologies to fully apply to the one-card system of Southwest University of Science and Technology.Using the relevant data mining algorithm to dig into the hidden data knowledge in a cartoon database is an essential part of our country's smart campus construction.At the same time,it also lays a foundation for the more mature data analysis system of colleges and universities.Through the data mining technology is more and more mature,we can dig out all kinds of valuable information,for the university management to fully understand the teachers and students in the school situation,make more conducive to the school management decisions.