Extraction of Consumer Emotion Using Diary Data on Purchasing Behavior
Extraction of Consumer Emotion Using Diary Data on Purchasing Behavior
复制标题
使用购买行为日记数据提取消费者情绪
DOI：
10.1007/978-3-031-10461-9_41
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Namatame Takashi
中科院分区：
文献类型：
--
作者：
Kitajima Yuzuki;Nakao Shunta;Otake Kohei;Namatame Takashi
近年来，随着消费者行为的多样化，超市和其他零售店在营销方面的竞争越来越激烈。在日本超市，主要的营销策略之一是活动营销。为了衡量未来公司之间的差异化，有必要推荐适合个人消费者需求的产品。理解消费者需求的方法之一是使用客户的声音或评论帖子进行自然语言处理。通过使用自然语言处理，可以估计导致消费者需求和决策的情绪。在这项研究中，我们关注消费者的情绪，目的是使用关于饮食习惯的购买行为的日记数据来估计复杂的情绪。具体地说，我们通过在现有的日语情感表达词典中添加日语版维基百科的文档，创建了一个新的日语情感表达词典。然后，我们使用词典从日记数据中估计每一篇文章的情绪。此外，我们还利用受试者对估计的情绪进行了问卷调查，并对结果的准确性进行了评估。
In recent years, supermarkets and other retail stores have become increasingly competitive in their marketing as consumer behavior has diversified. In Japanese supermarkets, one of the major marketing strategies is event marketing. To measure differentiation among companies in the future, it is necessary to recommend products that are suitable for the needs of individual consumers. One of the methods used to understand consumer needs is Natural Language Processing using customers’ voices or review posts. By using Natural Language Processing, it is possible to estimate emotions that lead to consumers’ needs and decision-making. In this study, we focus on consumers’ emotions and aim to estimate complex emotions using diary data on purchasing behavior about dietary habits. Specifically, we created a new Japanese emotional expression dictionary by adding documents of the Japanese version of Wikipedia to the existing Japanese emotional expression dictionary. We then estimated the emotion of each text from the diary data using the dictionary. Furthermore, we conducted a questionnaire survey using subjects for the estimated emotions and evaluated the accuracy of the results.