Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal Journal of Emerging Information Systems and Business Intelligence (JEISBI)
Darren Waluya Ardianto
Universitas Negeri Surabaya
Computer Science & IT Decision Sciences, Operations Research & Management Languange, Linguistic, Communication & Media Library & Information Science

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Combining Purchase Value and Product Preferences for Fashion Retail Customer Segmentation Darren Waluya Ardianto; I Kadek Dwi Nuryana
Journal of Emerging Information Systems and Business Intelligence (JEISBI) Vol. 7 No. 2 (2026): Vol. 07 Issue 02
Publisher : Universitas Negeri Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26740/jeisbi.v7i2.80883

Abstract

The rapid growth of transactions in the fashion retail industry produces large volumes of customer data whose behavioral patterns remain difficult to exploit when segmentation is performed manually and descriptively. Uniform mass marketing is particularly ineffective for premium fashion, where preserving brand exclusivity demands a more personal approach. This study develops a customer segmentation model that does not stop at grouping customers by transaction value, but also reveals the shopping-style preferences embedded within each segment, and presents the results through an analytical dashboard for business stakeholders. Recency, Frequency, and Monetary (RFM) parameters were derived from transaction data and transformed using logarithmic scaling and a Robust Scaler to handle skewed distributions and outliers. The parameters were then grouped with K-Means clustering, whose optimal number of clusters was determined through the Elbow Method, Silhouette Score, and Davies-Bouldin Index. Within each resulting cluster, Latent Dirichlet Allocation (LDA) was applied independently to uncover hidden purchase topics, which were interpreted as fashion-style personas. The findings show that 1,152 customers were divided into four tiers, with K=4 as the optimal configuration, supported by a Silhouette Score of 0.5952 and a Davies-Bouldin Index of 0.5644. LDA identified distinct style personas across tiers, with topic coherence shaped by each segment's vocabulary diversity. The resulting Streamlit dashboard passed all 45 functional test scenarios, confirming that integrating value-based and preference-based analysis yields richer customer profiles for targeted marketing.
Co-Authors I Kadek Dwi Nuryana, I Kadek Dwi