Bayesian Rough Set Model in Hybrid Kansei Engineering for Beverage Packaging Design

Abstract

Human evaluation have common shortage to capture vagueness and uncertainty while input multivariate data due to characteristics such relationship between packaging design attributes and customer requirement and perception about the package. In Kansei Engineering (KE), customer perceptions about a product tend to define the product value and this considered as whole of product attribute. The main objective of this work is to provide the designer with a robust formulation to make relationship between design element and customer perception. Then we proposed decision rules in order to get affective knowledge in bottle packaging design by using Bayesian Rough set method. This paper provided a construction of decision rules between design elements and customer perceptions such as slim shape of bottle body and bright colored of bottle cap to describe a modern bottling design. The result showed that Bayesian Rough Set model effectively extracted the decision rules of human evaluation data in designing beverage bottle from the intuition of customer perception. In conclusion our approach supported the design processes and eased the designer tasks