Design And Implementation Of Academic Early Warning System For University Students Based On Data Mining

Professionals training is the essential function of universities,and college students’ academic performance is an important indicator to measure students’ professional abilities.Undergraduate academic early warning is an important guarantee for colleges and universities to improve the quality of undergraduate education and the construction of learning style.First and foremost,based on the data mining technology,this thesis constructs an analysis model of students’ academic early warning,and uses the improved Apriori algorithm to mine association rules for students’ curriculum data.At the same time,based on the above research results,we successfully designed and implemented the college students’ academic early warning system.The system can not only provide students with academic early warning,but also provide useful data support and decision support for college teaching management.It is expected to have positive application and promotion value in the field of academic early warning in colleges and universities.The main contents of this thesis are summarized as follows:1.Taking the academic performance of undergraduates in the recent three years of the information school of a university as the data source,according to the division of semesters,the academic data set is constructed with statistical thought in view of the prominent problems such as the structure of the data source is cluttered and there are many missing values.In view of the low accuracy and accuracy of the algorithm of a single classification model,based on the Stacking framework,which is based on XGBoost,Light GBM,Random Forest and GBDT algorithms as the base learner,and a decision tree as a meta-learner,is constructed to mine the key features that affect undergraduate students’ studies from the data source.The experimental results show that the accuracy and accuracy of AWStacking model are 96% and 91%respectively,which is about 4% higher than that of single algorithm model.2.In order to facilitate the tracking of students’ professional course learning changes,based on the improved Apriori algorithm,taking the degree of interest,support and confidence as the measurement indicators,the general course,practical course,professional course and other course categories are taken into consideration,and the correlation analysis of student course data is carried out,so as to capture the characteristics of academic learning patterns of students of different majors and grades in different course categories from the multi-dimensional perspective of majors and grades,so as to provide data support for the undergraduate college’s student management.3.Based on the academic early warning model and the improved Apriori algorithm,a undergraduate academic early warning system is designed and developed based on the B/S architecture.It carries out multi-scenario and multi-angle academic early warning analysis from the three levels of students,departments and colleges,so that undergraduate students can grasp the academic details.Furthermore,the teaching management personnel may have a macro grasp of the overall academic situation of all grades and majors,in order to provide intellectual support and decision-making support for the management departments at all levels of the college.