FP-Growth in Discovery of Customer Patterns

摘要

The paper describes a knowledge discovery platform and a novel process for finding association rules based on the algorithm FP-Growth and its variants. Built software solution has been optimized in terms of memory usage and computation time as well as the impact of all modifications made to the whole process of rules discovery The process of rule discovery is illustrated on a real database containing transactions of customers of the e-shop Delicatessen Alma24.