Abstract
The Ganto website, ganto.co, is one source of information for students, lecturers, and the public. Every visit to the Ganto website is recorded on the server log. If left unchecked, the server log will only fulfill server storage. Therefore, actions need to be taken such as implementing Web usage mining. Web usage mining, using k-means clustering, is one way to ensure the satisfaction of users who access the Ganto website, by knowing the tendency of user access through server log analysis. First Cluster, September 2017. Second Cluster, October 2017, November 2017, December 2017, February 2018, March 2018, and April 2018. While the third cluster, July 2017, August 2017, January 2018, May 2018, and June 2018. Results of analysis , Berita is the most frequently viewed rubric, the second is Artikel, and the third is E-Paper. While the other rubrics (Info Kampus, Sastra Budaya, Ganto TV, and Ganto Foto) are not too often viewed by users or visitors of the Ganto website.
Keywords: Web usage mining, clustering, k-means, Ganto