Handling numeric behavioral attributes in actionable behavioral rules mining

Abstract

Actionable behavioral rules mining is a new problem of data mining. The produced rules can provide the user explicit suggestions of actions to influence the behaviors of the entity in concern with satisfactory utility to the user. When mining such rules, it is assumed that all the behavioral attributes are categorical, while numerical attributes have been discretized in advance. However, this assumption will hinder the performance of the algorithms for mining such rules. In this paper, to handle this problem, a numeric-behavioral-attributes-based problem definitions and a corresponding mining algorithm are proposed. The experimental results strongly suggest the validity and the superiority of our approach.