Abstract:In the current digital era, e-commerce has become one of the main pillars of global trade. With the ever-increasing amount of transaction and user activity data, e-commerce companies are faced with the challenge of understanding and managing diverse customer segments more effectively. This paper discusses the application of clustering algorithms for e-commerce customer segmentation based on purchasing data and user activity. The aim of this research is to identify homogeneous customer groups to support more targeted marketing strategies and increase customer retention. The problem faced is how to process big data originating from user transactions and activities on e-commerce platforms, as well as how to identify patterns that are useful for customer segmentation. The data used in this research includes purchase history, frequency of visits, length of time spent on the site, and interactions with certain products. The solution method applied in this research is the clustering algorithm, especially K-Means and DBSCAN. K-Means is used to group data into a predetermined number of clusters based on the Euclidean distance between data points. Meanwhile, DBSCAN is used to identify clusters with high density and separate data that is considered noise or outliers. Data preprocessing is carried out to clean and normalize the data before being applied to the clustering algorithm. Validation of clustering results is carried out using metrics such as Silhouette Score and Davies-Bouldin Index. The research results show that by applying the clustering algorithm, customers can be grouped into several segments that have similar characteristics. For example, we found groups of customers with high purchase frequency but low transaction value, as well as other groups with high transaction value but low purchase frequency. This information is very useful for companies to design more effective marketing strategies, such as special offers for customers with high transaction values or loyalty programs for customers with high purchasing frequency. The conclusion of this research is that clustering algorithms can be a very effective tool in e-commerce customer segmentation, allowing companies to understand customer behavior patterns and develop more targeted and effective marketing strategies. Thus, implementing this method is expected to improve business performance and overall customer satisfaction.

Implementation of the Fuzzy C-Means Clustering Method in Grouping Provinces in Indonesia based on the Types of Goods Sold in E-commerce Businesses in 2022

COMPARATIVE STUDY OF FUZZY C-MEANS AND K-MEANS ALGORITHM FOR GROUPING CUSTOMER POTENTIAL IN BRAND LIMBACK

Grouping of Retail Items by Using K-Means Clustering

Mapping the Wuling vehicle market with K-Means Clustering: An effective digital marketing strategy

Penerapan Algoritma Clustering untuk Segmentasi Pelanggan E-commerce berdasarkan Data Pembelian dan Aktivitas

Warehouse Location Optimization with Clustering Analysis to Minimize Shipping Costs in Indonesia’s E-Commerce Case

Automatic Categorization of Multi Marketplace FMCGs Products using TF-IDF and PCA Features

Pastry Sales Data Clustering with K-Means Clustering Approach for Product Grouping

Data Mining Modeling in Clustering Car Products Sales Data in the Automotive Industry in Indonesia

Utilizing Clustering Methods for Categorizing Delivery Requirements Based on Analysis of E-Commerce Product Data

ANALISIS CLUSTER DENGAN MENGGUNAKAN K-MEANS UNTUK PENGELOMPOKKAN ONLINE CUSTOMER REVIEWS (OCR) PADA ONLINE MARKETPLACE

Analysis Characteristics of Car Sales In E-Commerce Data Using Clustering Model

The application of K-means clustering for province clustering in Indonesia of the risk of the COVID-19 pandemic based on COVID-19 data

Optimization of Product Placement on E-commerce Platforms with K-Means Clustering to Improve User Experience

CLUSTERING OF CLOTHING SALES DATA AT TOP STORE USING K-MEANS METHOD

The impact of COVID-19 pandemic on non-cyclical consumer companies based on their series of financial performance using fuzzy c-means clustering

Data Mining on Women's Clothing Sales in Market Places with the K-Means Clustering Algorithm

Indonesian pharmacy retailer segmentation using recency frequency monetary-location model and ant K-means algorithm

A TWO-STEP CLUSTER FOR CLASSIFYING PROVINCES IN INDONESIA BASED ON ENVIRONMENTAL QUALITY

Performance of K-Means and DBSCAN Algorithm in Clustering Gross Regional Domestic Product

Penerapan Metode K-Means Clustering dalam Pengelompokan Data Penjualan Barang Pada PT. Swasti Tunggal Mandiri