【正文】 s fault description. Rulebased reasoning is used to guide the reuse of checkpoint solutions. Data mining for decision support Information, such as the best selling machines, the customers of a particular machine, a parison of sales among different machines, and the performance of different service engineers are highly desirable for the management team. 、 Data mining process Fig. 3 depicts the data mining process for extracting hidden knowledge from large databases. The process focuses on finding interesting patterns that can be interpreted as useful knowledge. It consists of seven steps. 、 Establishing the mining goals This involves the understanding of the customer service support process, its database, and the administrative procedures of the pany. A number of mining goals were identified: _ marketing: identify the machine models with poor sales and determine possible reasons。s machine, related manuals, and spare parts that may be required to carry out the repair. Such a process is inconvenient. At the end of each service cycle, a customer service report is used to record the new problem and the proposed remedies or suggestions taken to rectify it. This database is used for billing purposes, as well as for maintaining a corporate knowledge base. The service centre stores the customer service report in the database. Apart from maintaining a knowledge base on mon faults and its remedies, the customer service database also stores data on sales, employees, customers and service reports. These data are not only used for daytoday management operations, but help the pany in decision making on job assignment and promotion of service engineers, and marketing, manufacturing, and maintenance of different machine models. The customer service database serves as a repository of invaluable information and knowledge that can be utilized to assist the customer service department in supporting its activities. The objective of this paper is to discuss how to apply data mining techniques to extract knowledge from the customer service database to support two types of activities: decision support and machine fault diagnosis. The work was carried out as a collaborative work between a multinational pany and the School of Applied Science, Nan yang Technological University, Singapore. The pany manufactures and supplies insertion and surface mount machines for use mainly in the electronics industry. In traditional help desk service centres, service engineers provide a worldwide customer support service through the use of longdistance telephone calls. Such a mode of support is found to be inefficient, ineffective and generally results in high costs, long service cycles, and poor quality of service. With the advent of the Inter technology, it is possible to deliver customer service support over the World Wide Web. This paper describes a Webbased intelligent fault diagnosis system, known as WebService, to support customer service over the Web. In the WebService system, a hybrid casebased reasoning (CBR) and artificial neural work (ANN) approach is adopted as the intelligent technique for machine fault diagnosis. Instead of using traditional CBR technique for indexing, retrieval and adaptation, the hybrid CBR–ANN approach integrates ANN with the CBR cycle to extract knowledge from service records of the customer service database and subsequently recall the appropriate service records using this knowledge during the retrieval phase. Data mining Data mining, also known as knowledge discovery in databases (KDD), is a rapidly emerging field. This technology is motivated by the need of new techniques to help analyze, understand or even visualize the huge amounts of stored data gathered from business and scientific applications. It is the process of discovering interesting knowledge, such as patterns, associations, changes, anomalies and significant structures from large amounts of data stored in databases, data warehouses, or other information repositories. It can be used to help panies to make better decision to stay petitive in the marketplace. The major data mining functions that are developed in mercial and research munities include summarization, association, classification, prediction and clustering. These functions can be implemented using a variety of technologies, such as database oriented techniques, machine learning and statistical techniques. Recently, a number of data mining applications and prototypes have been developed for a variety of domains including marketing, banking, finance, manufacturing and health care. In addition, data mining has also been applied to other types of data such as timeseries