高级检索

AI生成评论总结对用户参与评论行为的影响——基于DID的准自然实验研究

The Impact of AI-Generated Review Summaries on Users' Review Participation Behavior —— A Quasi-Natural Experiment Study Based on DID

  • 摘要: 随着生成式人工智能的快速发展,各平台陆续推出AI生成在线评论总结(AIGORS)缓解消费者信息过载。但AIGORS在提升信息获取效率的同时,对消费者参与评论行为的影响尚未明确。研究以TripAdvisor和携程国际版(Ctrip)644家酒店的评论数据为研究对象,采用双重差分法(DID)考察AIGORS引入后用户评论数量增速的变化,探究其对评论行为的作用,并进一步分析评论累计量、产品评分的调节机制。研究发现:AIGORS显著抑制消费者评论行为,该效应随评论累计量增加、产品评分升高而强化,低评论累计量、低产品评分场景下抑制作用不明显。研究结论完善了在线评论与AI生成内容交叉领域的理论体系,为平台优化AIGORS设计、平衡信息效率与用户参与度提供了实践指导。

     

    Abstract: With the rapid development of Generative Artificial Intelligence (AIGC), various platforms have successively launched AI-Generated Online Review Summaries (AIGORS) to alleviate information overload among consumers. However, while AIGORS improves the efficiency of information acquisition, its impact on consumers' review participation behavior remains unclear. Taking review data of 644 hotels on TripAdvisor and Ctrip International as the research object, this study adopts the Difference-in-Differences (DID) method to examine the changes in the growth rate of user review volume after the introduction of AIGORS, explore its effect on review behavior, and further analyze the moderating mechanisms of cumulative review volume and product rating. The study finds that AIGORS significantly inhibits consumers' review behavior, and this effect is strengthened with the increase in cumulative review volume and product rating, while the inhibitory effect is not significant in scenarios with low cumulative review volume and low product rating. The conclusions of this study enrich the theoretical system in the interdisciplinary field of online reviews and AI-generated content, and provide practical guidance for platforms to optimize AIGORS design and balance information efficiency with user participation.

     

/

返回文章
返回