|
|
Applying fuzzy-association-rule techniques and artificial neural networks to customer Kansei knowledge mining |
SHI Fu-Qian1,2, SUN Shou-Qian2, XU Jiang2 |
1. Computer Department, Wenzhou Medical College, Wenzhou 325000, China;
2. Modern Industrial Design Institutie, Zhejiang University, Hangzhou 310027, China |
|
|
Abstract Customers’ Kansei information on product geometry features is obtained through Web survey system. Customers’ Kansei evaluation is in fuzzy expression and then multi-dimension fuzzy-association-rule technology is applied. Andfinally, high-frequency sets in which the geometry features of products are highly associated with customers’ Kansei information are generated. Furthermore, the learning ability of BP neural network are utilized to train, predict and integrate association rules from different time. And then customers’ Kansei knowledge mining is implemented, which provides new ideas for product aid design and enterprise decision-making.
|
Published: 28 August 2007
|
基于模糊关联与BP网络的客户感性知识挖掘
通过建立Web问卷调查系统获取用户对产品造型特征的感性反映信息,并对用户感性评价信息予以模糊表征,进行多维模糊关联法则挖掘,进而产生客户感性信息与产品造型特征关联规则高频项目集。利用BP神经网络的学习能力对不同时段关联规则进行训练、预测和整合,从而实现客户感性知识挖掘,为产品设计辅助与企划决策支持提供新思路。
关键词:
模糊关联法则,
倒传递类神经网络,
产品设计
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
|
Shared |
|
|
|
|
|
Discussed |
|
|
|
|