Reserach Express@NCKU - Articles Digest

This research presents a total new approach to integrate the multi-disciplines of Kaisei Engineering and Type 1 Numerical Theory into Neural Networks. The new approach can successfully transform users’ perception into product elements design. In this research, product image has been extended on product form to include product color. Hence, the most important design problems as the follows have been successfully solved ：

1. What kind of product form is suitable for a certain color?
2. Are all product forms suitable for any color?
3. Which has a greater influence on product image to the users?
4. Does color affect product image and thus emotional feeling of the users?
5. Is there an optimal combination of form and color for a given design concept of product image?

This research uses mobile phones as an experimental example. Three experimental stages were conducted to prove the new approach. In the first stage, the concept of Kansei Engineering were used to extract the representative experimental samples, including product form samples, product color samples, and image word pairs of product image. Forty-five product design experts were used as subjects, which have been separate into three groups. In this stage, 4 representative product form samples have been extracted from 70 mobile phones in the market, 72 representative product color samples extracted from The color system of Commission International de l’Eclairage (CIE) and 3 representative image word pairs of product image samples extracted from more than 100 image word pairs dada base. Hence there are 864(4×72×3) representative mobile phones samples.

In the second stage, Quantitative Theory Type I analysis were used to examine the degree of relative importance of product form and product color, including 4 product elements (form and 3 color attributes), on product image. In this research, 4 independent variables (X₁, X₂, X₃, and X₄) are used to represent one form (phone style) and 3 color attributes (Lightness, Chroma, and Hue). And, 3 dependent variables (Y₁, Y₂, and Y₃) are used as the average of product image evaluations on the 3 image word pairs respectively. The result of Quantitative Theory Type I analysis is given in Table 1. The category grades shown in Table 1 indicate the preference degree of the user perception on the each category of independent variables. If the grade is negative, the user perception leans towards the left image. On the contrary, the positive grade indicates that the user perception favors the right image.

Table 1 The results of preference degree of Quantitative Theory Type I
(Source：Hsin-Hsi Lai, Yang-Cheng Lin, Chung Hsing Yeh, Chien-Hung Wei, User-oriented design for the optimal combination on product design, International Journal of Production Economics, vol. 100, April, 2006.pp.253-267)

In the final stage, Neural Netwoks techniques were used to evaluate the optimal combination of product image. In order to examine whether the NN model is an effective technique for determining the optimal combination on product design for matching a desirable product image, two NN models were developed for the training process, called mixed-NN model and single-NN model respectively in the experimental study. Table 2 shows the results of performance evaluation of the mixed-NN and single-NN models. The results obtained from Table 2 can be used to help product designers to work out the optimal combination of product form and product color for a particular design concept represented by a product image word pair. Figure 1 shows the optimal combinations of product form and product color for a
particular set of product images.

Table 2 results of performance evaluation of the mixed-NN and single-NN models.
(Source：Hsin-Hsi Lai, Yang-Cheng Lin, Chung Hsing Yeh, Chien-Hung Wei, User-oriented design for the optimal combination on product design, International Journal of Production Economics, vol. 100, April, 2006.pp.253-267)

Fig. 1. The optimal combinations of product form and product color represented by the S-C, H-R, and L-F values of 2.
(Source：Hsin-Hsi Lai, Yang-Cheng Lin, Chung Hsing Yeh, Chien-Hung Wei, User-oriented design for the optimal combination on product design, International Journal of Production Economics, vol. 100, April, 2006.pp.253-267)

The results of this research provide useful insights in designing form and color of a product for enhancing the image of the product. These design-supporting results would help product designers work out the optimal combination of product form and product color on product image. Although mobile phones are chosen as an illustration of the new approach, this innovative approach can also be applied to other products with various design elements.