Design And Implementation Of Student Automatic Grouping System Based On Feature Clustering

With the rapid development of data mining technology,many industries benefit from the application of big data analysis.Education is in a new era of reform and development."Data promotes campus,and analysis revolutionizes education" is making the educational practice more personalized and humanizing.Educational domain under the background of big data applications starts relatively late compared with other industries.Collaborative learning,which is supported by the computer technology,has attracted more and more attention from researchers around the world as the cross point of computer and education.Collaborative learning research focuses on finding a more scientific,rational and effective strategy for collaborative learning groups.To solve the problem,this thesis intends to find a grouping method fits the psychological indicators of learners,and is highly individualized and easy to achieve.The research contents of this thesis mainly include the following aspects:Firstly,a grouping model based on feature clustering is proposed.By comparing the advantages and disadvantages of traditional grouping,combining with the principle and method of educational psychology,the thesis puts forward a cluster grouping model based on the characteristics of learning style and academic record.Secondly,the clustering analysis of experimental data is completed.The experimental data is extracted through online test and the data is normalized according to its validity and completeness.The characteristics of learners are analyzed by means of K-means algorithm.The study finds that learners' learning style has obvious clustering tendency in clustering algorithm,and it is refined into four typical feature classifications,which lays the foundation for the subsequent grouping strategy.Thirdly,the design of student grouping system based on feature clustering is implemented.According to the basic needs of learners' group,this thesis puts forward the implementation of the strategy respectively for homogeneous grouping,heterogeneous grouping,comprehensive grouping and expounds the implementation process.Eventually,the experimental indexes examine the homogeneous grouping,heterogeneous grouping,and comprehensive grouping strategy has the positive promotion on the learners' learning.Under the real teaching scene,the experimental data indicates that the automatic grouping system in the heterogeneous grouping shows a better evaluation than traditionalgrouping.It prove that the grouping strategy of this subject can effectively improve the teaching effect.