ARHyMST anonymized retail transaction dataset

This record contains the anonymized retail transaction dataset used in the study “From Redundant Association Rules to Product Networks: A Scalable Confidence-Improvement Pruning and Maximum Spanning Tree Approach”. The dataset is provided as a compressed ZIP archive. After extraction, the input files are located in the data/ folder, which should be kept unchanged because the analysis scripts expect this directory structure.

The dataset contains anonymized retail transaction records used to construct approximately 2.64 million basket transactions from 14.9 million product records aggregated into 46 product groups. Basket transactions were constructed by grouping purchases made by the same customer on the same date and through the same sales channel.

The corresponding source code is available at: https://github.com/sasapgd/arhymst-market-basket-analysis