感性工学是研究产品中消费者自身情感需求的一个重要方法论，探索双向感性工学以建立"设计属性-感性属性-感性偏好"之间的关系对于设计者和消费者有着重要的感性决策支持作用。设计属性和感性属性的权重信息对于构建属性之间的关系有着重要影响；双向感性工学中存在着主观评价信息和属性之间关系的不确定性。本项目从质量管理中的KANO及QFD模型和标签语义理论出发来开展双向感性工学研究。首先，基于标签语义理论构建不确定性定量感性KANO和感性QFD模型以得到感性属性的分类等级及其权重和设计属性的权重；其次，在包含感性偏好的感性评价实验和属性权重信息的基础之上，提出基于标签语义理论的语言决策树模型以构建属性之间的不确定性关系；最后，整合上述模型形成双向感性决策支持方案及系统并将之应用到智能手机和酒店服务两种类型的产品之中。研究成果能够为设计者和消费者提供直观、透明及容易理解的知识表示框架和更好的感性决策支持。

Kansei Engineering (KE) has been developed as an important methodology to translate the technology of a consumer's feeling and image for a product into the design elements of the product. It is important to study two-way KE so as to build the relationships among ``Design attribute-Kansei attribute-Kansei preference" for providing Kansei decision support to designers and consumers. In the two-way KE, the weighting information of different attributes plays an import role in building the relationships among them. Furthermore, there exists high-level uncertainty in the two-way KE: uncertainty of Kansei information and uncertianty of the relationships between different attributes. Towards this end, this study will deal with the two-way KE from the perspectives of label semantics in computing with words and the KANO, QFD models in quality management. To do so, firstly, uncertain Kansei KANO and uncertain Kansei QFD models are constructed in order to obtain the hierarchy class of different Kansei attributes with their weighting information and the weighting information of design attributes, respetively. Secondly, with these weighting information and the Kansei experiment involving Kansei preference in mind, fuzzy linguistic decision tree model has been built based on the label semantics to caputre the uncertain relationships among different attributes. Finally, the above-mentioned models have been integrated to formulate Kansei decision support solutions/system for designers and consumers; the Kansei decision support system is applied to two types of products: smart phones and hotel service. As a conclusion, this study can provide direct, transparent and easily understood models with a high-level knowledge representation framework and better Kansei decision support to designers and consumers.