【正文】 ver 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, spatial, telemunications, web, and multimedia data. In general, the data mining process, and the data mining technique and function to be applied depend very much on the application domain and the nature of the data available. Customer service support Service records (or reports) are currently defined and stored in the customer service database. Each service record consists of customer account information and service details, which contain two types of information: faultcondition and checkpoint information. The former contains the service engineer39。s description of the machine fault, while the later indicates the suggested actions or services to be carried out to repair the machine, based on the actual faultcondition given by the customer. Checkpoint information contains checkpoint group name, and checkpoint description, with priority and an optional help file. The checkpoint group name is used to specify a list of group checkpoints. Each checkpoint is associated with a priority that determines the sequence in which it can be exercised and a help that gives visual details on how to carry out the checkpoint. An example of faultcondition and checkpoint information for a service record is given in Fig. 2. In addition, the customer service database also stores data related to sales, customers, and employees: six major tables are defined in the customer service database for this. Two, namely, MACHINE_FAULT and CHECKPOINT, are used to store the knowledge base on mon machine faultconditions and their checkpoints. These are unstructured textual data. The remaining four tables are used to store information on customers (CUSTOMER), employees (EMPLOYEE), sales (MACHINE) and maintenance (SERVICE_REPORT). These four store only the structured data. There are over 70000 service records. Since each of the faultconditions has several checkpoints, there are over 50000 checkpoints. Information on over 4000 employees, 500 customers, 300 different machine models and 10000 sales transactions are also stored. 、 Mining structured data A list of most popular data mining tools available mercially or in public domain is given in the KD Nuggets website. These tools can mine the structured data of sales, maintenance, and particulars of employees and customers in the customer service database. It is interesting to see that a number of tools support multiple approaches。 ., more than one data mining techniques. For example, Darwin from Thinking Machine Corp. supports neural works, regression tree (CART), kmeans algorithm, and case based reasoning for classification, prediction, and clustering functions. There are also some tools that only aim at a special data mining function. This provides 175。edibility。 the users can select different data mining tools for their