• 论文 •    

基于粗糙集的感性知识关联规则挖掘研究

石夫乾,孙守迁,徐江   

  1. 浙江大学 计算机学院现代工业设计研究所,浙江杭州310027
  • 出版日期:2008-02-15 发布日期:2008-02-25

Association rule mining of Kansei knowledge based on rough set

SHI Fu-qian,SUN Shou-qian,XU Jiang   

  1. Institute of Contemporary Industry Design, School of Computer Science & Technology, Zhejiang University, Hangzhou 310027,China
  • Online:2008-02-15 Published:2008-02-25

摘要: 为有效获取概念设计中的感性信息,提出了一种基于粗糙集的感性知识关联规则挖掘方法。首先,在用户进行感性调查与产品造型特征组合评价的基础上,运用统计学方法,生成一个由感性词汇索引的决策表;其次,通过粗糙集理论对相关属性进行约简,以提取对相应感性评价贡献较大的造型特征;再次,进一步构造出基于粗糙集的关联运算法则及规则合并,以缩小决策表规模。最终得到了关键特征对感性描述的强关联规则集,并以此引导概念设计定位和开发。实例证明,该方法在手机产品的感性知识关联规则挖掘中得到了较好的应用。

关键词: 感性工学, 粗糙集理论, 关联规则挖掘

Abstract: To acquire Kansei information effectively in the conceptual design, a new method of association rule mining for Kansei knowledge based on rough set theory was proposed. Firstly, with combinatorial evaluation of the product form features, a decision table indexed by Kansei words was generated by the statistical analysis on user′s Kansei questionnaire. And then, relevant attributes were simplified by rough set theory so as to extract the key form features which contributed largely to the corresponding Kansei evaluation. Association computation rules and their mergence were constructed to reduce the scale of the decision table. Finally, strongassociation-rule-sets were generated to guide the orientation and development of the conceptual design. The proposed method was implemented successfully in the cell phone design case.

Key words: Kansei engineering, rough set theory, association rule mining

中图分类号: