A study on gender-kansei of three-dimensional geometric shapes

Rajapakse, R.P.C. Janaka; Jayasinghe, H.S.; Miyata, Kazunori; Marasinghe, Ashu; Tokuyama, Yoshimasa

doi:10.1504/ijbm.2012.049741

Cited by 4 publications

(1 citation statement)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Comparative cross-cultural study of craftsmanship assessment for vehicle interior design found no significant differences between two Nationality groups of the subjects that were all males mechanical engineering student [21]. Also, the study by Rajapakse [22] which have found no significant differences in response to the product shape in male and female. This issue is raised in the widely recognized research on product form relation with consumer response written by Bloch [20].…”

Section: B Ideal Product Design Conceptmentioning

confidence: 89%

Application of Kansei Engineering and Data Mining in Developing an Ingenious Product Co-design System

Sakornsathien¹,

Sinthupinyo²,

Anuntavoranich³

2019

IJMLC

View full text Add to dashboard Cite

This research intends to be a part in developing an automatic system that shows the potential of designing a product form by co-designing with the user in order to make it suitable for each consumer by applying Kansei Engineering Technique. However, the operation of KE needs data collection from specific target group in order to devise a formula and interpret the results when applying to study the same target group, or else the accuracy will be reduced. In order to eliminate the drawbacks, Data Mining techniques will be applied with KE system in order to use as an ingenious product co-design system.The style and preference of each user will be used as a categorizing factor clustering the database into groups with K-means technique, in this research. Each classifying cluster will use its own database in the system processing in order to gain a set of design elements precisely from the system. Decision Tree classification technique is selected for the study. For the model validation, we apply the cross-validation as an unbiased model performance evaluation. In order to build the KE system, the sentiments in this research are indicated by 5 pair-words constituting the system data. The result shows that, this method offers more accurate prediction of design elements comparing to the method which undivided users.Index Terms-Data mining, machine learning, Kansei engineering, product development.

show abstract

Section: B Ideal Product Design Conceptmentioning

confidence: 89%