Implementation of the K-Mean Algorithm to Determine the Level of Student Satisfaction with the Online Learning Uhamka System (OLU)

Abstract

The Covid-19 pandemic has had a negative impact on humans not only in health but also in the economy, social and education. Schools and colleges were closed so that learning that was originally carried out face-to-face was shifted to long distance learning (LdL). LdL implementation can be carried out synchronously and asynchronously. There are obstacles in learning using the Online Learning Uhamka (OLU), namely the effectiveness of using the Online Learning Uhamka (OLU) and the application of the k-means algorithm to determine the level of student satisfaction with the Online Learning Uhamka (OLU) system and the reliability of the k-means algorithm in clustering .One technique to measure the level of satisfaction is to use clustering techniques. The advantage of the clustering technique is that it is easy to adapt, imply and execute and is commonly used in various fields. One of the clustering techniques that is often used is the k-means algorithm. There are 2 clusters used in the k-means algorithm. Clustering results using the Kmeans algorithm showed that 309 respondents belonged to cluster 1, namely satisfied, and 94 respondents belonged to cluster 2, namely dissatisfied. The indicators used to assess satisfaction are usability, content quality, interaction quality. Of the three assessment indicators that have the lowest score is the interaction quality indicator with the centroid value in cluster 1, namely 19.33980583 and the centroid value in cluster 2, namely 14.08510638. The results of the Kmeans algorithm reliability test by calculating the Davies Bouldin index value are good enough in clustering data. The Davies Bouldin index value is 0.3806830859.