Affiliations: [a] School of Arts and Design, Nanchang University, Nanchang, China
| [b] Design School, Loughborough University, Loughborough, United Kingdom
| [c] Faculty of Technology, Art and Design, Oslo Metropolitan University, Oslo, Norway
Correspondence:
[*]
Corresponding author. Xinhui Kang, School of Arts and Design, Nanchang University, No. 999 Xuefu Avenue, Nanchang 330031, Jiangxi, China. E-mail: nbukxh@163.com.
Abstract: Customer satisfaction is an important indicator of user preferences for products which directly influence customers’ purchase intentions. It is also an essential Kansei factor for enterprises to successfully develop products. Therefore, the objective of this study is to apply the fuzzy weighted association rule mining (FWARM) approach to extract the significant association between customer satisfaction and product form features, thus providing specific parameter guidelines for the enterprise’s business decisions. In previous research, the fuzzy association rule mining (FARM) approach has proved to be a promising way forwards. However, the absence of consideration of the weight of items is always criticized as creating an uninteresting rule with high frequency and low importance. Therefore, in this study the fuzzy Delphi method (FDM) is used to calculate the weight of each item in the early stage of data mining, and filters out items with a high degree of consensus. Then, the FARM method extracts the fuzzy weighted association rule. Taking the household exercise bike as an example, the authors find that handlebar, LCD screen, rack, main outline and pedestal are vital item features. Subsequently, the method is used to identify 14 rules to inform the development of the exercise bike’s form to achieve high customer satisfaction; valuable knowledge support is provided for manufacturers and designers in the initial stage of new product development, potentially improving customer satisfaction and reducing the risk of product failure.
Keywords: Customer satisfaction, Kansei engineering, Fuzzy Delphi method, Fuzzy weighted association rule mining, Exercise bike
