Combination Grouping Techniques and Association Rules For Marketing Analysis based Customer Segmentation

Changes in people's transaction behavior using the internet resulted exponential growth of e-commerce. With digital shopping transactions, it is difficult to predict customer segments and patterns traditional mathematical models. Timely identification emerging trends from large volumes data plays a major role business processes decision making. This different previous research works that apply RFM model based on K-Means Clustering find potential customers as an ingredient determining marketing targets. In this study, clustering technique approach proposed classify which evaluated Davies Bouldin, Calinski Harabasz Silhouette methods determine optimal number clusters, then results are used Apriori algorithm goods often purchased together. Based test Clustering, Spectral Gaussian Mixture Model techniques produced 5 clusters with 76% more accurate method than other two so was determined RMF model, grouping were concurrent product purchases by expected be useful future management.