Employing rough sets and association rule mining in KANSEI knowledge extraction

Abstract

KANSEI Engineering (KE) is a method for translating feelings and impressions into product parameters and the objective of KANSEI Engineering is to study the relationship between product forms and KANSEI images. It is most important to extract critical form features of the product relative to specific KANSEI adjectives through a WEB-based KANSEI information system. In this paper, critical form features and KANSEI adjectives were defined as condition attributes and decision attributes respectively, which were formalized as two objects in Decision Table (DT). Then, the Semantic Differential (SD), which measures the connotative meaning of concepts, was applied to evaluate form features of the product through a KANSEI questionnaire system. The evaluation record from an individual’s transaction data was reserved if its frequency was higher than the given threshold. Some form features were deleted by using an attribute reduction algorithm based on Rough Sets Theory (RST). Furthermore, the size of the DT was reduced by using a rule-joining operation. A strong association rule set which describes the relationship between the critical form features and the corresponding KANSEI adjectives was subsequently generated. A case study of a mobile phone design was presented to demonstrate the effectiveness of the proposed method by comparing it with other non-linear data mining methods in KANSEI Engineering.