基于潜在语义分析和感性工学的用户需求匹配  

User needs matching based on latent semantic analysis and kansei engineering

作　　者：刘征宏 谢庆生[1] 李少波[1,2] 林丽[2] 

机构地区：[1]贵州大学现代制造技术教育部重点实验室,贵州贵阳550003 [2]贵州大学机械工程学院,贵州贵阳550025

出　　处：《浙江大学学报（工学版）》2016年第2期224-233,共10页Journal of Zhejiang University：Engineering Science

基　　金：国家自然科学基金资助项目(51475097;51465007);国家"十一五"科技支撑计划资助项目(2014BAH05F01)

摘　　要：为了更加精确地满足用户需求并提高设计精度,结合潜在语义分析(LSA)和感性工学(KE),提出基于潜在语义分析和感性工学(LSA-KE)的用户需求匹配方法.通过构建RKE模型,获得感性词汇对和设计元素权重;使用RKE模型构建LSA-KE语义空间和计算设计方案的感性值,通过用户调查获得用户需求感性值;使用感性工学和LSA-KE语义空间匹配用户需求和设计方案,并与匹配结果进行对比.以机床造型设计为案例对该方法进行描述,结果表明该方法能明显提高用户匹配精度.A new LSA-KE approach was presental based on latent semantic analysis（LSA）in order to meet user needs more accurate and improve product design efficiency.LSA was used match terms to documents for document indexing,and kansei engineering（KE）was used to match user needs to designs.A RKE model was created to obtain the KE weights for all design elements and all Kansei word pairs.The KE model was used to establish a LSA-KE semantic space and calculate Kansei values for new designs,determine Kansei values for user needs through survey users simultaneously.The KE and the LSA-KE semantic space were used to match Kansei values for user needs to Kansei values for new designs.Then the two results were compared.Application process and procedure of the proposed method were described by a case of numerical control machine tools form design.The LSA-KE approach was used to match Kansei values for user needs to numerical control machine tools designs for each user.The case study results reveal that the LSA-KE approach can significantly improve matching accuracy.

关 键 词：用户需求 潜在语义分析 感性工学 数控机床造型设计 

分 类 号：TH166[机械工程—机械制造及自动化]