An analysis method of dynamic requirement change in product design

Abstract

Previous studies conducted customer requirement (CR) mining through the Kano classification approach to optimize product design, without considering the opinion changes of manufacturers and customers over time. To fill this gap, this paper discusses the methods for requirement change in product design from both customer and manufacturer perspectives. A Customer–Manufacturer-Kano (CM-Kano) model is proposed to analyze the ever-changing opinions of customers and manufacturers to provide improvement strategies for product design. Accordingly, two mechanisms have been proposed to provide support for product improvement, (1) a classifier of CM-Kano is used to dynamically classify the attributes based on opinion changes, and (2) an improvability index is introduced as a ranking factor for CRs to support product design and improvement. First, the authors measure the importance weights and sentiment orientations of attribute words extracted from online texts and capture the opinion changes of customers and manufacturers over time. Next, the effects of opinion changes are calculated based on a Radial Basis Function Neural Network (RBF-NN). On this basis, the CM-Kano model is proposed to identify the categorization of product attributes. An illustrative case study was implemented on vehicle improvement design, which demonstrates that the CM-Kano model can effectively improve product design.