Prioritizing Association-Rules for Product Recommendation: A Case Study In E- Retailing

E-commerce web sites enable businesses to utilize recommendations that help to increase cross-sales by providing relevant products. The traditional Market Basket Analysis relies on finding products often purchased together in prior basket data. In such cases, selecting the most appropriate offer requires evaluation of association rules through various strategies for rule selection. With an emphasis on the recency of transactions, this study presents a slight modification in the Apriori algorithm, which is widely used in association rule discovery. Weighted rule mining techniques help to prioritize transactions regarding attributes, including the recency, customer, or purchase total. Our study demonstrates a case study where the itemset support calculation with a weighted approach favors more recent transactions. The findings indicate notable decreases in completion time, while prioritizing more recent transactions with the aim to obtain time-sensitive and potentially more useful recommendations.