Design And Implementation Of University Academic Warning System Based On Hadoop

In recent years,the country’s emphasis on college education has been increasing,the expansion of colleges and universities has also increased,and the number of college students has also increased.This has created many problems and challenges.Among them,the academic problems of college students are particularly important,such as the decline in the quality of students,The number of students dropping out of courses is increasing,and the dropout rate of students is increasing.The decline in academic quality has had a negative impact on universities,individuals and families.Therefore,it is urgent to establish an academic early warning system for universities,which can provide planned and targeted assistance to students,so that students can complete their studies smoothly and become the right choice.A useful member of society.However,due to the large number of disciplines in universities and the intricacies of students’ academic data,if you want to dig out valuable information from these large amounts of data,the traditional stand-alone computing model obviously cannot meet the growing demand.In this context,this thesis designs and implements a college academic early warning system,combined with the Hadoop distributed Map Reduce computing framework,to achieve the prediction and early warning of student performance,to help students improve their studies,and has theoretical and practical significance for the training of college talents.The design ideas of the paper are as follows:Useing Hadoop platform technology to build a distributed data processing platform to provide a theoretical basis for parallel computing of algorithms.Secondly,using FP-Growth association rule analysis and K-Means algorithm,combined with the advantages of the Hadoop platform,apply it to academic analysis to provide school leaders and teachers with targeted suggestions and solutions.The parallelized K-Means algorithm is used to mine the distribution between the scores of each subject,and based on the results of the clustering,specific and reasonable suggestions are made for different types of students;the parallelized FP-Growth algorithm is used to conduct data mining on the academic courses of the students who fail to take the subject.Obtaining the internal relationship between disciplines is conducive to making a reasonable training plan.Finally,combined with the SSH framework,an academic early warning system was built through the integration of the presentation layer,business logic layer and data persistence layer.The system is composed of four parts: early warning data management,early warning information management,early warning data mining,and system basic information management.On the one hand,it realizes the academic information and early warning information of students at different stages,which helps teachers and students keep abreast of the academic overview;On the one hand,the system can perform early warning analysis of student performance.For example,by digging out the relationship between disciplines and disciplines with a high dropout rate,it is convenient for colleges and universities to formulate effective training programs in a timely manner to provide teaching for colleges and universities.Theoretical support to promote the school’s smart campus construction.