Achmad Reza Mardika, Student ID Number: 06.2014.1.06373
Dian Puspita Hapsari, Advisor ID Number: 153076 Weny M. Rahmawati, Advisor ID Number: 153101 Department of Informatics Engineering, Faculty of Information Technology reza.mardika.rm@gmail.com
ABSTRACT
Online retail means sell-buy transaction from customers to sellers through
internet. Some recent years, the development of online retail grows so rapid that many online retail companies are interested to have data mining to cluster their customers based on the similar characteristics and identify their differences for the sake of marketing strategy development. The aim of this research is to investigate customers mapping in regard to customer relationship management. The researcher employed RFM Analysis for customers assessment based on recency, frequency, and monetary, as well as Fuzzy C-Means algorithm for clustering process. Fuzzy C-Means is a clustering algorithm based on member levels. The result of clustering process obtained six clusters based on the six characteristics of customers. The quality of clustering results was then tested by SSE (Sum Squared error) algorithm. The six clusters tested by SSE got validity by 58071.3. This validity was then compared with different parameters of c and w. The most optimum cluster was resulted from parameters of 4 clusters and 2 weighting exponents as they got the smallest cluster validity by 22566.5.