Establishing method of campus student relation networks

Abstract

The invention discloses an establishing method of campus student relation networks. The establishing method includes: collecting campus behavior data of students, establishing access time sequences of locations of the students and calculating co-occurrence frequencies of specific locations of every two students; through multiple shuffling of the access time sequences of the locations of the students, and establishing empty models to calculate the co-occurrence frequencies of every two students in a calculation random situation; through comparative analysis of the empty models, determining true co-occurrence frequencies of the locations of the students and generating the relation networks of the locations among the students; through a linear weighting method, aggregating the co-occurrence relation networks of all the locations. The co-occurrence frequencies of the students are considerably calculated from the access time sequences of the students, random co-occurrence situations of the students are filtered through the empty models, valuable data of the relation networks among the students is extracted, and an important role is played for student management staff in strengthening student management and understanding student characteristics and social-economic status.

Description

technical field [0001] The invention relates to network data mining technology, in particular to a method for constructing a campus student relationship network. Background technique [0002] Higher education administration is student-centered and aims to prepare them to be co-occurrors and leaders in every aspect of their lives. An important factor in higher education management is the prediction of academic performance, which provides reference for educators. For example, if educators knew about students' academic performance in advance, they could intervene in student instruction in a timely manner, thereby preventing academic failures most likely. This is very important for education management, because academic failure largely affects students' graduation, job hunting and future development. There are methods to predict a student's academic performance based on information from different sources, such as self-reports by test takers, behavioral data on smartphone senso...

AI Technical Summary

Problems solved by technology

However, there is still little research on inferring campus interpersonal relationships and constructing a complete campus social ecosystem, especially based on large-scale behavioral data of many students.

Images