Data mining for customer service support

In traditional customer service support of a manufacturing environment, a customer service database usually stores two types of service information: (1) unstructured customer service reports record machine problems and its remedial actions and (2) structured data on sales, employees, and customers for day-to-day management operations. This paper investigates how to apply data mining techniques to extract knowledge from the database to support two kinds of customer service activities: decision support and machine fault diagnosis. A data mining process, based on the data mining tool DBMiner, was investigated to provide structured management data for decision support. In addition, a data mining technique that integrates neural network, case-based reasoning, and rule-based reasoning is proposed; it would search the unstructured customer service records for machine fault diagnosis. The proposed technique has been implemented to support intelligent fault diagnosis over the World Wide Web. (C) 2000 Elsevier Science B.V. All rights reserved.