Research On The K-Means Clustering Algorithm Based On Hadoop In The Evaluation Of College Students' Comprehensive Quality

With the expansion of college enrollment,the number of college students is rapidly increasing,but while the demand gap of high quality talents in society get bigger and bigger.So,how to properly evaluate students,and how to improve the quality of personnel training? These questions need to be considered and solved.In this paper,a mass of students' data are obtained from information systems and using data mining techniques we can get a deeper level of valuable information from them,to provide services for the university personnel training.The main research contents can be summarized as follows:First,get to know about the current situation of students' comprehensive quality evaluation in colleges through literature and researches,and then analyze existing problems.In consideration of the attention today's society is paying to high-qualified talents,build a more scientific and reasonable students' overall quality evaluation index system.After using AHP,a high reliable method which can reduce uncertainty to a large degree,and consulting a number of experts,the weight values of all indicators have been finalized.Then,using Hadoop platform,which has powerful parallel computing and mess data analyzing and processing abilities,parallels the K-Means clustering algorithm to improve the efficiency of the processing,and make clustering analyses of real data.The results play an important role in optimizing the overall development of students and higher education teaching programs and improving personnel training programs.Finally,a college students' comprehensive quality evaluation system based on the evaluation model has been developed.It can make a more scientific,comprehensive,accurate and objective information evaluation and the K-Means clustering analysis function has also been implemented,so that the student managers can be relieved from the huge workload.It can also reduce the risk of error and improve the evaluation efficiency.Furthermore,by showing the results very intuitively,the system can assist with management decision-making.