Communications of the ACM
Knowledge engineering and management: the CommonKADS methodology
Knowledge engineering and management: the CommonKADS methodology
Data mining: concepts and techniques
Data mining: concepts and techniques
Product Configuration Frameworks-A Survey
IEEE Intelligent Systems
A constructive method for multivariate function approximation by multilayer perceptrons
IEEE Transactions on Neural Networks
| Hi-index | 12.05 |
Mass confusion is a metaphor of the burdens for customers as a result of attractive but probably overloaded options in mass customization interaction process. This perceived risk of mass customization has been revealed and discussed in last decade and most research concentrates on the personalization of the interaction behaviour between front-end roles, i.e. customers and salesmen. Although the research in the front-end helps to improve customers' satisfaction, the risk of mass confusion cannot be well leveraged by isolated personalization in the front-end: the intensive involvement of customers and salesmen also brings challenges to their patience, qualification and the responsiveness of configuration activities. Differing from most existing research, this research tries to propagate the effort of personalization from front-end roles to back-end configuration knowledge by proposing a kind of personalized configuration rules (PCRs). To simultaneously alleviate the difficulties in PCR acquisition, representation and deduction, both symbolic and connectionist methodologies are adopted to establish dual formulations for PCR development. Although these two methodologies are theoretically disparate, approaches based on specific rule extraction technique are proposed to coordinate the dual formulations and make them complementary in application. A case study of the proposed methodology for bicycle configuration is presented.