Research On Personalized Recommendation And Student Team Matching Based On Student Profile

With the rapid development of higher education and the increasing demand for talents in the society,innovation and entrepreneurship education has gradually become a research hotspot in current higher education.With the continuous development of innovation and entrepreneurship education,personalized recommendation is more and more valued in innovation and entrepreneurship education;at the same time,subject competition as an important form of innovation and entrepreneurship education,its scale is growing,and most competitions require multiple students build a team,and the demand for team formation is increasing.Therefore,in the process of promoting innovation and entrepreneurship education,how to portray students through various data,so as to better serve students,carry out more in-depth study recommendations and subject competition team matching and other related research has become an urgent problem to be solved.This article mainly builds a campus big data platform,collects and analyzes multiple data such as theoretical learning and practical learning of students to construct student profile,and on this basis,personalized test question recommendation and subject competition team matching.The research contents are as follows:First,research and construct a student profile based on the campus big data platform.Through the analysis of the existing systems on campus,combined with the behavioral characteristics of students in an innovative and entrepreneurial environment,a series of software systems based on IPv6(Internet Protocol Version 6)are developed to enrich student data,thereby establishing a campus big data platform.Through the processing,analysis and modeling of the data collected by the campus big data platform,a student profile is established.Then,research and implement personalized test question recommendation based on student profile.In view of the problem that students' learning status is difficult to judge,this paper proposes a fuzzy cognitive diagnosis model based on student profile to make reasonable judgments on students' learning status attributes;make predictions about the answer situation;meanwhile,based on the utility theory,design a utilitybased test question recommendation method to recommend students to answer highutility test questions.The experimental results show that the F-DINA(FuzzyDeterministic-Input,Noise-And)method proposed in this paper has a higher accuracy than the traditional DINA(Deterministic-Input,Noise-And)method and DINO(Deterministic-Input,Noise-Or)method in predicting the results of test questions;in the recommendation system,the F-DINA method has better performance than the DINA method and DINO method,and the average utility of the recommended test questions of the F-DINA method is also higher than that of the DINA method and DINO method,which can effectively improve the students' income in answering test questions.Finally,the team matching of subject competition based on student profile is researched and realized.In view of the difficult and unreasonable current situation of student team formation in the current subject competition,this paper proposes a team matching model for the subject competition based on student profile.This model models the student and the competition,uses the utility theory to take the competition team's utility as the overall benefit,and uses the intelligent algorithm to maximize the utility of the student team,so that the team and the student can obtain higher profits.The experimental results show that the CTB(Competition team build)method proposed in this paper has better performance than the traditional team matching method in the simulation environment or the real environment,and improves the overall effectiveness and ability of students and teams.At the same time,it can also narrow the individual differences between students and teams and avoid the polarization of traditional methods.