Reseach On Mapping Methodology For Affective Design

Affective product design aims at incorporating customers’ affective needs into designvariables of a new product so as to optimize customers’ affective satisfaction. The coreissue in the field of affective design is to identify the mapping relationships betweenKansei words and design elements. Although some methods have been proposed to buildaffective mapping models, there are still many problems to be solved. This dissertationaims to improve this affective mapping methodology system. The reseach contents andconclusions are as follows:(1) To consider the heterogeneity in customer group, a robust design method inquality engineering is introduced to design a Kansei product which can satisfy customergroup’s mental imagery and decrease the fluctuation of Kansei reactions to the productamong different customers as far as possible. The concrete precedure is as follows: first,the concept of Kansei distance is proposed to quantify the Kansei quality of a product;then, regarding the Kansei parameters of the product as control factors, the highlyindividualized characteristics among customers as noise factors, and the Kansei distanceas response variable, the approach of response modeling based on single arrays isproposed to obtain the robust settings of a product. A case study of the Kansei design for amobile profile was conducted to illustrate and validate the proposed method.(2) Ordinal probit regression (OPR) is introduced into the field of affective design todiscover mapping knowledge from design elements to customer affect, and a comparativestudy is always recommended between OPR and ordinal logistic regression (OLR) foravailable data to choose a better fitted model. The discovered mapping relations couldfacilitate the handling of affective information and assist the designer to make trade-offdecisions. A case study of cell phone design was conducted. Four generic affectivedimensions and six key product attributes of the cell phone were identified. OPR and OLRwere applied successively to reveal the quantitative relations from design elements tocustomer affect. The two models were compared in terms of goodness of fit, predictability,global significance, significance of coefficients and significance level of proportionalodds assumption. The findings show that OLR model is superior to OPR model to fit thecollected data. (3) For the affective mining procedure, a method which could refine association ruleseffectively is presented: at first, a set of raw rules are generated by specifying low valuesfor the support and confidence thresholds; then, these raw rules are evaluated to refine themost meaningful rules, in this step, a series of concrete sub-steps is presented. A casestudy Volvo truck cab design was conducted to illustrate the proposed method.