<![CDATA[The development of information technology has encouraged local governments to improve the quality of public services through digital media, one of which is the official government website. However, the effectiveness of local government websites still varies and not all of them provide an optimal user experience. This study aims to analyze the effectiveness of local government websites in South Sumatra Province using the K-Means clustering algorithm with the help of the RapidMiner Studio application. The data used is secondary data from SimilarWeb, covering four main indicators, namely total visits, average visit duration, pages per visit, and bounce rate. The results of the analysis show that local government websites are divided into three main clusters, namely effective, less effective, and ineffective. Effective websites have high visit rates, long visit durations, and low bounce rates, while ineffective websites have the lowest performance on all indicators. The variables of average visit duration and bounce rate proved to be the most dominant factors in determining website effectiveness. These findings are expected to serve as a basis for local governments in evaluating and optimizing their websites so that digital public services are more efficient, interactive, and responsive to the needs of the community.]]>
Copyrights © 2025
