A STUDY FOR CAPTURING DEEP IMPRESSION BY UNITING NATURAL LANGUAGE PROCESSING AND NETWORK CONSTRUCTION TECHNIQUES

在设计产品时，设计师需要创造出符合用户情感感受的产品。这项研究假设存在某种情感“印象”，这是用户从产品中获得的，预计会影响用户的偏好。在本研究中，我们关注的是深层印象，它被定义为与对产品的深层感受相关的印象，并隐藏在表面印象之下。为了捕捉深刻印象的本质，我们开发了一种构建“虚拟印象网络”的方法，其中涉及“结构”和“不明确印象”的概念，使用语义网络。这篇论文尤其旨在理解人们形成偏好印象的方式。我们的研究结果表明，可以使用虚拟印象网络的网络理论中的几个指标来解释“喜欢”和“不喜欢”的感觉之间的差异。“喜欢”的印象形成过程与“不喜欢”的印象形成过程在深刻印象层面上有所不同。

When designing products, designers need to create products that are congruent with the emotional feelings of users. This study assumes that there are some sorts of emotional "impressions", which a user derives from a product that is expected to affect user preference. In this study, we focus on deep impressions, which are defined as the impressions that are related to the deep feelings towards a product and lie under surface impressions. In order to capture the nature of deep impressions, we developed a method for constructing "virtual impression networks," which involve the notions of "structure" and "inexplicit impressions," using a semantic network. This paper, in particular, aims at understanding the manner in which people form impressions of preference. Our results indicated that it is possible to explain the difference between feelings of "like" and "dislike" using several indicators in the network theory of virtual impression networks. The process of forming the impressions of "like" is shown to differ from that of "dislike" at a deep impression level.