Online Course Popularity: An In-depth Analysis with K-Means on Udemy Platform

Abstract

Online courses have become increasingly popular in the digital age, offering flexibility, access to materials and enhanced collaboration. However, online course providers face challenges in understanding user preferences to develop effective marketing strategies. This study uses the K-Means clustering algorithm to analyze the popularity of online courses on the Udemy platform. The analysis is based on a Kaggle dataset containing course attributes such as enrollment numbers, reviews, and others. The elbow method was used to determine the optimal number of clusters, with results showing that K=4 is optimal. The clustering quality was assessed using the Davies-Bouldin Index (DBI) and Silhouette Score, yielding a DBI value of 0.65171 and a Silhouette Score of 0.6269, indicating good cluster separation. The study reveals four distinct clusters of courses, each with varying popularity and characteristics. This study can help course providers create more targeted marketing strategies, improving user engagement and satisfaction based on the specific traits of each course group.