Kansei Clustering Using Fuzzy and Grey Relation Algorithms

Kansei engineering (KE) has been widely used as a consumer-oriented technique to better understand dcustomers' emotional responses and further translate them into the design elements of a product. Kansei clustering plays a vital role in the implementation of KE. Conventional KE approaches rely heavily on the intuition of designers to cluster the Kansei adjectives; however, such classifications may be inconsistent with customer opinions. This paper presents a Kansei clustering approach using fuzzy and grey relation algorithms. In the proposed algorithms, topology-based grey relational analysis (TGRA) is used as an indicator function to derive a set of relational grades for constructing a fuzzy proximity matrix. Fuzzy relation-based clustering associated with a cluster validation index (CVI) is the nused to classify collected Kansei adjectives into an appropriate set of product attributes. An empirical study is also presented to demonstrate the implementation process of the proposed approach.