Design And Research Of Student Information Management Platform Based On Campus Big Data

With the rapid development of information construction in modern society,the education management in universities should also focus on the smart campus.The popularity of campus card and the accumulation of records of major digital platforms jointly create a big data environment on campus.Campus data stores students’multidimensional data,which also has the characteristics of large-scale,high-speed,multi-type.How to excavate valuable information from these massive data,assist university administrators to better understand students’ behavior characteristics and mentality,timely pay attention to abnormal data and make early warning,is an important goal in the process of university information construction.The project integrates data mining technology and traditional statistical analysis technology in the process of university data management,integrates multi-dimensional campus original data,adopts corresponding research schemes for different research objectives,implements the concept of meticulous and precise management of students in Campus,builds a student information management platform,and helps the construction of smart campus.The main contents are as follows:(1)Collect students’ multi-source campus data,and integrate,clean and desensitize the original data.(2)Construct three-dimensional student portrait.By building a four layer student label system,and inputting the model labels into the two-stage outlier detection clustering model,namely lof local outlier detection algorithm and improved k-means clustering algorithm,we can generate all kinds of three-dimensional student portraits,and find the groups that need to pay attention to and excellent students.(3)This paper studies the identification model of students from poor families.Combined with student behavior data and data mining technology,this paper analyzes the correlation between student behavior characteristics and family economic status.Based on the multi-dimensional feature collection training fusion judgment model,scientific and objective generate funding judgment results,respond to the national accurate funding policy.(4)Assist the supervision and management of campus.This paper analyzes the data of outlier students,and pays attention to the students with closed social contact,irregular life data and low academic performance from different perspectives.Combined with the analysis results of traditional SCL-90 and questionnaire,this paper discusses the quantifiable indicators related to students’ emotion fluctuation.Timely positioning and warning of abnormal data of student groups,so as to facilitate managers’ offline attention.