K-Means to Determine the e-commerce Sales Model in Indonesia

Abstract

Sales of e-commerce in Indonesia are currently experiencing a pretty good increase by starting to switch sales that are done traditionally or offline to online sales using the use of internet media to facilitate trade transactions between sellers and buyers can be done anywhere and anytime. In this study using the K-means clustering method algorithm. The data source used in this study is based on sales using e-commerce based on provinces in Indonesia in 2018. The criteria used in the calculation are divided into 3 clusters, namely seller, reseller and dropship. Centroid data for cluster 1 = 68.30,33,58,4.95 namely Aceh, Riau, West Kalimantan, East Kalimantan, and Maluku. Centroid data for cluster 2 = 78.58,22,25.4,22 i.e. North Sumatra, West Sumatra, Jambi, Bangka Belitung Islands, D.K.I Jakarta, Central Java, East Java, Banten, Bali, Central Kalimantan, South Kalimantan, South Sulawesi and Papua and Centroid data for cluster 3 = 86.36,13.34,4.38 namely South Sumatra, Bengkulu, Lampung, Riau Islands, West Java, D.I Jogyakarta, West Nusa Tenggara, East Nusa Tenggara, North Kalimantan, Central Sulawesi, Gorontalo, West Sulawesi, North Maluku and West Papua.