关键词: 产品语意关联, 分类树, 回归树, 产品特征

Abstract: To achieve a user preference value as high as possible, a design approach to obtain Product Semantics (PS) and Product Characteristics (PC) was proposed based on the Product Semantics Relationship (PSR) mechanism. Ordinary PC and PS were obtained by employing many technologies, including semantic differential method, factor analysis, cluster analysis and shape analysis. With the proposed method, regression tree technology was particularly used to achieve the user's most satisfactory PSR. This PSRrelated PC was also specifically obtained by classification tree technology. The presented method was applied to the design of sun glasses, and a scheme's preference value reached as high as 2.89. Owing to its high preference value, this design approach could be expected to gain widespread application in the fashion product design.

Key words: product semantics relationship, classification tree, regression tree, product characteristics