The Students Group Detection Based on the Learning Styles and Clustering Algorithms

Abstract

The approach to automatically student groups detection based on Honey and Mumford's questionnaire and index of learning styles questionnaire with the help of clustering methods are proposed in the paper. The methodology of our research consists of the following stages: 1) the evaluation optimal number of clusters and clustering students data based on Honey and Mumford's questionnaire and Ward.D2 method realised in R-library NbClust; 2) the evaluation optimal number of clusters and clustering students data based on index of learning style questionnaire and k-Means method; 3) forming the clusters of students that are in one cluster based on both learning styles and description of the student clusters taking into account similarities on learning style preferences and similar interests in the social network.