Customer Value Evaluation Based on Rough Set with Information Gain and Generate Decision Tree
Ming-Chang Lee *
National Kaohsiung University of Applied Sciences (Taiwan), 415 Chien Kung Road, Kaohsiung, Taiwan.
*Author to whom correspondence should be addressed.
Abstract
Aims: Customer value can be used to segment customers, and to determine which customers should be the focus of marketing efforts and dollars. Attributes are reduced by rough set theory, redundant attributes are removed and the core attributes are gained. Information gain is used in the classification of objectives and the analysis of this information. The aim of this paper provides customer value evaluation model using rough sets theory and classification algorithm (generated decision tree) by attribute information gain. The contribute of this paper is useful for developing data pre-processing, removing redundant attributes, and extracting decision rule from an instance of the customer value analysis.
Design: The purpose approach is (1) Data preparation (2) Finding an optimal reduces using rough set theory (3) Generating rule tree (4) Rule extraction module.
Methodology: This paper solves customer value evaluation problem using Rough set theory, entropy theory, Generate-decision- tree, and Information gain.
Results: The results are that The proposed model is useful for developing data preprocessing, removing redundant attributes, and extracting decision rule from an instance of the customer value analysis.
Keywords: Rough set, information entropy, information gain, customer value, generate decision tree