基于KANO-IPA模型的用户满意度分析——以人机交互体验的智能个人助理产品的在线评论为例  

Users Satisfaction Analysis Based on KANO-IPA Model--Taking Online Review of Intelligent Personal Assistant Product in Human-computer Interaction as An Example

作　　者：蒋昀霖 夏志杰[1] JIANG Yunlin;XIA Zhijie(School of Management,Shanghai University of Engineering Science,Shanghai 201620,China)

机构地区：[1]上海工程技术大学管理学院,上海201620

出　　处：《科技与经济》2024年第4期106-110,共5页Science & Technology and Economy

基　　金：国家社会科学基金一般项目——“大数据支持下网络谣言智慧治理机制及运行策略研究”(项目编号:21BGL243,项目负责人:夏志杰)成果之一;上海市社科规划一般课题——“大数据时代伪健康信息传播特征及多主体协同干预研究”(项目编号:2020BGL005,项目负责人:夏志杰)成果之一。

摘　　要：通过在线评论这一用户表达需求的重要渠道,可以识别和评估产品或服务对用户满意度的影响,进而为企业提供改进产品或服务的启示。将KANO-IPA模型和数据挖掘技术相结合,综合挖掘用户满意度和产品属性的重要性。首先,对用户的在线评论进行数据挖掘,利用BERTopic进行主题建模,以此构建属性-情感词典;其次,将产品属性映射到KANO-IPA模型,对属性进行分类,并确定用户对各属性的偏好和优先级;最后,以人机交互中的智能个人助理产品为例,分析产品对用户需求的优先级。结果表明,针对不同用户需求提出差异性优化建议,对产品的改进与提升具有重要意义。Online reviews are important channels for users to express their needs,and it can also identify and evaluate the impact of products or services on user satisfaction,and then provide inspiration for enterprises to improve them.This paper proposed the KANO-IPA model,which can comprehensively explore the importance of user satisfaction and product attributes.First,data mining of online reviews was carried out,and topic modeling was conducted by using BERTopic to build attribute-emotion dictionary.Second,product attributes are mapped to the KANO-IPA model,attributes are classified,and users preferences and priorities for each attribute are determined.Finally,taking the intelligent personal assistant product in human-computer interaction as an example to analyze,the priority of the product to the user s needs is analyzed.The research shows that it is of great significance to put forward optimization suggestions according to different user needs for product improvement and promotion.

关 键 词：在线评论 KANO-IPA模型 数据挖掘 满意度分析 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术] F49[自动化与计算机技术—计算机科学与技术] F713.55[经济管理—产业经济]