<![CDATA[The use of e-wallets has grown rapidly along with the increasing demand for fast and conven-ient digital financial transactions. This development requires service providers to better under-stand user behavior through transaction pattern analysis. This study aims to cluster e-wallet users based on their transaction patterns using the K-Means algorithm. The data analyzed in-clude several key variables, namely transaction frequency, transaction value, transaction type, and transaction time. The K-Means method is applied to group users with similar transaction characteristics through data normalization and optimal cluster determination. The results in-dicate that e-wallet users can be divided into several distinct segments with significantly dif-ferent transaction patterns. Each cluster represents specific user behavior characteristics, such as high-frequency users, high-value transactions, or time-based usage patterns. In conclusion, this study contributes to the growing body of research on digital payment analytics by demonstrating the applicability of the K-Means clustering algorithm for transaction-based user segmentation. The findings provide empirical evidence that behavioral transaction data can be systematically structured into meaningful user groups through unsupervised learning techniques. This research extends prior studies by integrating clustering evaluation methods to ensure optimal segmentation results, thereby offering a methodological reference for future studies in fintech data analysis.]]>
Copyrights © 2026
