Abstract: The fitness issue of online clothing purchase has stimulated the development of a number of virtual fitting technology-based fitting software. In order to investigate the status quo of online virtual fitting terminal experience service, online user comments of the mainstream virtual fitting software in the current market were crawled using crawler technology from the perspective of user perception, and the virtual fitting service quality index system was established through sentiment orientation analysis and LDA theme modeling. The results are as follows: the existing virtual fitting service can basically meet consumers' fitting needs; the image of virtual fitting model, fitting style, system software upgrades and other services need to be improved; users are most concerned about 7 theme dimensions of system security, system design, response speed, service quality, fitting function, interactive quality and content information; Finally, based on the above analysis, a virtual fitting platform service quality evaluation index system with 7 first-level indexes and 25 second-level indexes was constructed in order to improve the online shopping service experience.

Key words: virtual fitting, service quality, evaluation index, LDA theme model

摘要： 线上服装购买的适体性问题,催生了一批以虚拟试衣技术为核心的试衣软件。为掌握线上虚拟试衣终端体验服务现状,从用户感知角度出发,利用爬虫技术,爬取目前市场上主流的虚拟试衣软件用户在线评论,通过情感倾向性分析和LDA主题建模,建立了虚拟试衣服务质量指标体系。结果发现:现有虚拟试衣服务基本满足消费者试衣需求;虚拟试衣模特形象、试衣款式、系统软件升级等服务内容亟待改进;用户最关注系统安全、系统设计、响应速度、服务品质、试衣功能、交互品质、内容信息7个主题维度服务;并依据上述分析构建了7个一级指标,25个二级指标的虚拟试衣平台服务质量评价指标体系,以提升线上购物服务体验。

关键词: 虚拟试衣, 服务质量, 评价指标, LDA主题模型