电商用户评价驱动的产品感性意象解构与外观转化设计研究——以家用3D打印机为例

Research on the Deconstruction of Product Perceptual Imagery and Appearance Transformation Design Driven by E-Commerce User Reviews—A Case Study of Household 3D Printers

摘要: 针对电商环境下产品迭代快、用户感性需求捕捉难的问题，本研究旨在探索一种从非结构化电商评论到结构化设计语言的量化转化路径。首先，采集电商平台家用3D打印机的多源数据，利用大语言模型(LLM)进行细粒度语义挖掘，提取产品属性与感性因子；其次，引入IPA-KANO模型对用户需求进行优先级映射，确定核心设计维度；随后，通过感性工学方法构建形态特征与意象评分的映射模型，利用机器学习算法预测最优外观方案。研究成功识别出影响家用3D打印机视觉偏好的核心要素，并推导出“科技智能”、“高端奢华”等多种意象驱动下的设计策略，通过设计实践产出了具有市场针对性的方案。本研究建立的转化框架实现了电商“小数据”向设计知识的有效迁移，不仅提升了感性设计的精准度，也为数智化背景下的工业设计流程提供了科学的决策支撑。

Abstract: Aiming to address the challenges of rapid product iteration and the difficulty of capturing users’ emotional needs in the e-commerce environment, this study explores a quantitative transformation pathway from unstructured e-commerce reviews to structured design language. Firstly, multi-source data on household 3D printers were collected from e-commerce platforms. A large language model (LLM) was employed for fine-grained semantic mining to extract product attributes and emotional factors. Secondly, the IPA-KANO model was introduced to map the priority of user demands, thereby identifying core design dimensions. Subsequently, a mapping model between morphological features and image ratings was constructed using Kansei engineering methods, with machine learning algorithms applied to predict optimal appearance solutions. The study successfully identified the core elements influencing the visual preferences for household 3D printers and derived design strategies driven by various images, such as “technological intelligence” and “high-end luxury”. Through design practice, targeted market-oriented solutions were developed. The transformation framework established in this research achieves an effective transfer of “small data” from e-commerce into design knowledge, enhancing the precision of emotional design and providing scientific decision-making support for industrial design processes in the context of digital intelligence.