Configuration Rules Acquisition for Product Extension Services Using Local Cluster Neural Network and RULEX Algorithm

Manufacturers are combining products and services to provide greater value to the customers. The bundling of physical products and product extension services (PESs) is the strategy adopted by manufacturers most frequently. To enhance customer value, the variety of PESs significantly increases to respond to different kinds of customer needs, which inevitably results in the configuration problem. In the systematic configuration problem, configuration rules acquisition is important to the effectiveness and efficiency of configuration solution. However, PESs configuration rules are hard to induced since there exists various domain knowledge in the new manufacturing paradigm. Thus, the authors propose an approach combining Local Cluster Neural Network and RULEX Algorithm to extract knowledge (i.e. rules) from historical data. A case study on copier PESs is illustrated to validate the approach.