Customer Product Choice Recommendation By Association Rules And Learning Models

Authors: Keerti Pal, Jayshree Boaddh, Rahul Patidar

Abstract: Online stores and apps attracts customer at various levels. So approaches need to be more effective by analyzing the behavior of visiting customer. Many of researcher has proposed different models of customer product recommendation system. This paper introduces a novel Customer Product Recommendation by Rules and Ensemble Model (CPRRESM) framework designed to enhance purchase prediction accuracy for small-scale retail stores with limited data resources. The proposed approach integrates Apriori-based association rule mining for pattern discovery, Z-score normalization for feature standardization, and a Gradient Boosting ensemble model for efficient learning. By combining rule-based insights with ensemble learning, CPRRESM effectively captures customer purchasing behavior and dynamic preferences.

DOI: http://doi.org/10.5281/zenodo.18107972