Statistical Visualization Methods for Tabular Data in Kansei Engineering
In Kansei engineering, several tabular data are crucial for Kansei engineering analyses. Design element table contains samples design variables and their variations. Kansei evaluation values are also made into a table. These tables have higher dimensional data space, because of large number of variables in both rows and columns. The aim of study is to develop a method of creating a graphical analysis that shows the relationship between a Kansei evaluation and a given set of design elements. Two methodologies are shown here. First is an attempt to visualize 2-factors evaluation experiment. This approach utilizes Local Regression Smoothing to show relationship between design elements and Kansei evaluation. Second attempt is visualizing larger design element table. In this approach, all samples are mapped on a two-dimensional plane according to a statistical analysis of the design elements. The coordinates of each sample in the map are solved by the Quantification Theory Type Ⅲ model using the computation method of correspondence analysis. Next, a three-dimensional contour map is created for the specific Kansei word on which the researcher or product designer wants to focus. Adding a Kansei evaluation value for each sample as a height value augments the map. Then a smooth contour that interpolates between the Kansei values of the samples is computed by a local regression method. The proposed methodology creates a three-dimensional contour map that helps researchers to recognize both linear and nonlinear relationship between a Kansei evaluation and the design variables.
Shigekazu Ishihara Keiko Ishihara Mitsuo Nagamachi
Hiroshima International University
国际会议
17th World Congress on Ergonomics(第十七届国际人类工效学大会)
北京
英文
1-8
2009-08-09（万方平台首次上网日期，不代表论文的发表时间）