CUSTOMER SEGMENTATION MENGGUNAKAN FUZZY C-MEANS CLUSTERING PADA E-COMMERCE HENZ COLLECTION

Abstract

Online shop "Henz Collection" is an onlineshop that sells products such as clothes, bags, and shoes. The main problem faced by the online shop are not yet have a website that specialized in selling the products that it sells itself and difficult to know which customers have loyalty to this shop online. Ecommerce application is built using Fuzzy C-Means. Fuzzy C-Means clustering is a technique in which the existence of each data point is determined by the degree of membership (partition). Value centroid k-means clustering taken from random numbers, so that when the centroid determination will be modified so that the clustering results have not changed, so that the customer data can be grouped consistently and data on the cluster have been no changes during the period that would diklaster not changed. The purpose of the application design e-commerce based website is to look at the criteria for customers who are entitled to a rebate / discount. Based on test results and data modules can be concluded that the application is running as expected. Discounting seen from the experimental formation of clusters that provide the best global value silhouette. Cluster in the period 2014 to have the degree of similarity of 60% of customer data and in the period 2015 to have the degree of similarity of 80% of customer data contained in these clusters with the data of the owner onlineshop discount.