Article ID: | iaor20127336 |
Volume: | 54 |
Issue: | 1 |
Start Page Number: | 142 |
End Page Number: | 152 |
Publication Date: | Dec 2012 |
Journal: | Decision Support Systems |
Authors: | Zhao Huimin, Zeng Daniel, Su Peng, Mao Wenji |
Keywords: | decision, datamining |
Many applications can benefit from constructing models to predict the behavior of an entity. However, such models do not provide the user with explicit knowledge that can be directly used to influence (restrain or encourage) behavior for the user's interest. Undoubtedly, the user often exactly needs such knowledge. This type of knowledge is called actionable knowledge. Actionability is a very important criterion measuring the interestingness of mined patterns. In this paper, to mine such knowledge, we take a first step toward formally defining a new class of data mining problem, named actionable behavioral rule mining. Our definition explicitly states the problem as a search problem in a framework of support and expected utility. We also propose two algorithms for mining such rules. Our experiment shows the validity of our approach, as well as the practical value of our defined problem.