Clustering Student's Satisfaction in Complex Adaptive Blended Learning with the Six Value System Using the K-Means Algorithm

Dadang Sudrajat,Raditya Danar Dana,Arif Rinaldi Dikananda,Nining Rahaningsih,Dian Ade Kurnia

doi:10.13189/ujer.2019.070920

Abstract

Complex adaptive blended learning with the six value system is expected to improve the quality of learning. It is also expected to improve high order thinking skills. But the level of student’s satisfaction in Complex adaptive blended learning with the six value system varies in results. This study aims to determine the grouping of student’s satisfaction in complex adaptive blended learning with the six value system. The object of the study was conducted in 3 (three) vocational high schools (SMK) in Cirebon City, West Java Province, Indonesia for digital simulation subjects with a purposive sample of 150 students. Data about student’s satisfaction is grouped using the K-Means algorithm with the optimization of generation method. Several stages are carried out in grouping student’s satisfaction, starting with randomly determining initial centroid values. The K-Means algorithm process ends if there is no change in centroid value between one iteration and another iteration. Furthermore, performance measurements are performed using the Cluster Distance Performance method. The results are obtained by the performance Vector with parameters K = 4. The average distance in the centroid is -0.107, the average distance in the center of cluster 0 is -0.109, the average distance in the center of cluster 1 is -0.109, the average distance in the center of cluster 2 of -0,100, and the average distance in the center of cluster 3 is -0,106, with Davies Bouldin index of -1,049.

Highlights

Blended learning is a combination of face-to-face learning in a traditional environment; on the other hand, there is an e-learning environment as information and communication technology that is likely to be expanded for communication and interaction distribution
The problem faced in this research is clustering or grouping students who are doing learning activities with the level of satisfaction of students consisting of very satisfied students, satisfied students, students are quite satisfied and students are less satisfied so we need a data clustering technique
The results of the clustering process obtained by the centroid table as shown in table 2 are as follows: Table 2

Summary

Introduction

Blended learning is a combination of face-to-face learning in a traditional environment; on the other hand, there is an e-learning environment as information and communication technology that is likely to be expanded for communication and interaction distribution. Teacher-centered learning is no longer relevant to the rapid development of information and communication technology so it needs to be modified [2] - [15]. One alternative learning model that can be used is complex adaptive blended learning [9]. Along with the development of information technology that is increasingly rapid and more complex, it is seen that the tendency of learning based on complex adaptive blended learning systems is increasingly complex. Complex adaptive system is a system that takes place despite changes in diverse individual components, because the interaction between these components is responsible for the survival of the system, and the system itself is involved in adaptation or learning [9]. The Complex Adaptive Blended Learning System states that among the components in the Complex Adaptive Blended Learning System, those involved in interaction in blended learning include teacher, learner, institution, technology, learning support and content [17][18][19]

Objectives

Methods

Results

Conclusion