Mining Consumer Services Based on User Preference with Associative and Process Mining

Abstract

A customer makes a decision based on sentiments when using services. These sentiments drive the selections in services that are correlated with each other. It is hard to grasp the sentiments of consumers when many combinations of options are available. This paper proposed a method to extract preferred services based on consumer behavior with strong correlation values between options in a service. The approach is by mining the association rules that correspond to the business process model. We first formalized a problem for extracting the preferred service model. Then, we proposed an extraction method by pruning the associative rules with more substantial relation using Cook’s distance and visualizing consumer behavior with a process model. Finally, we illustrated the approach of the proposed method and showed that we could extract preference with a higher correlation value compared to the conventional method.