An Intelligent Customer Retention System
This paper proposes an intelligent system for handling the customer retention task, which is getting important due to keen competition among companies in many modern industries. Taking wireless telecommunication industry as a target of research, our system first learns an optimized churn predictive model from a historical services database by the decision tree-based technique to support the prediction of defection probability of customers. We then construct a retention policy model which maps clusters of churn attributes to retention policies structured in a retention ontology. The retention policy model supports automatic proposing of suitable retention policies to retain a possible churner provided that he or she is a valuable subscriber. Our experiment shows the learned churn predictive model has around 85% of correctness in tenfold cross-validation. And a preliminary test on proposing suitable package plans shows the retention policy model works equally well as a commercial website. The fact that our system can automatically propose proper retention policies for possible churners according to their specific characteristics is new and important in customer retention study.
KeywordsAssociation Rule Customer Relationship Management Association Rule Mining Learning Mode Package Code
Unable to display preview. Download preview PDF.
- 1.Oceansblue: Accurate Customer Churn Prediction with Information Inventor, available at: http://www.oceansblue.co.uk/resources/iichurn1.pdf
- 10.Chunghwa Telecom: Best Package Plan Calculation, available at: http://www.cht.com.tw
- 11.Stanford Medical Informatics: The Protégé Ontology Editor and Knowledge Acquisition System, available at: http://protege.stanford.edu
- 12.Quinlan, J.R.: Induction of Decision Tree. Machine Learning 1, 81–106 (1983)Google Scholar
- 13.Quinlan, J.R.: C4.5: Programs for Machine Learning. Morgan Kaufmann, San Mateo (1993)Google Scholar
- 15.Agrawal, R., Srikant, R.: Fast Algorithms for Mining Association Rules. In: Proc. of the 20th International Conference on Very Large Databases, pp. 487–499 (1994)Google Scholar