Paper Info
Title
Detecting the change of customer behavior based on decision tree analysis
Abstract
Understanding and adapting to changes in customer behavior is an important aspect for survival in a continuously changing environment. This paper develops a methodology based on decision tree analysis to detect the change in classified customer segments automatically between two data sets collected over time. We first define three types of changes as the emerging pattern, the unexpected change and the added/perished rule. Then, similarity and difference measures are developed for rule matching to detect all types of change. Finally, the degree of change is developed to evaluate the amount of change. Our suggested methodology based on decision tree analysis in the change detection problem can be used in more structured situations in which the manager has a specific research question and it also detects the change of classification criteria in a dynamically changing environment. A Korean Internet shopping mall case is evaluated to represent the performance of our suggested methodology, and practical business implications for this methodology are also provided. We believe that the change detection problem and the suggested methodology will become increasingly important as more data mining applications are implemented.
Year
DOI
Venue
2005
10.1111/j.1468-0394.2005.00310.x
EXPERT SYSTEMS
Keywords
Field
DocType
data mining,decision tree,change analysis,Internet shopping mall
Data mining,Decision tree,Data set,Change detection,Research question,Computer science,Consumer behaviour,Rule matching,Artificial intelligence,Internet shopping,Machine learning,Incremental decision tree
Journal
Volume
Issue
ISSN
22.0
4.0
0266-4720
Citations 
PageRank 
References 
12
0.75
20
Authors
4
Name
Order
Citations
PageRank
Jae Kyeong Kim1101152.32
Hee Seok Song21578.75
Tae Seong Kim3120.75
Hyea Kyeong Kim437412.77