%0 Journal Article
%T 考虑消费者体验型产品特征偏好的评论排序研究<br>Review Ranking Considering Consumers’ Preferences for Experiential Product Features
%A 吕钰菁
%A 陈磊
%J Management Science and Engineering
%P 1-12
%@ 2167-6658
%D 2024
%I Hans Publishing
%R 10.12677/MSE.2023.131001
%X 在线评论能够降低消费者在体验型产品购买决策中的感知风险，但评论信息过载对高效的搜索构成了挑战。由于体验型产品具有个性化的特点，消费者对评论信息的需求因产品特征偏好而异，这一问题在现有研究中尚未得到充分阐述。研究通过LDA2Vec、LCR等方法，构建消费者类模型及基于类的评论有用性预测模型，从而实现评论个性化排序。实验结果证实了个性化排序模型在提高评论感知有用性上的有效性，其在评论相关性、完整性、信息诊断性、消费者满意度等方面都显著优于有用性投票排序机制。本研究为如何识别体验型产品特征，并通过消费者体验型产品特征偏好解决体验型产品评论排序提供了具体的方法，更为通过评论个性化排序缓解评论信息过载问题提供了理论见解、模型成果和经验证据。<br />
Online reviews serve as a potent tool in diminishing the perceived risk consumers’ face when making purchasing decisions regarding experiential products. However, the challenge lies in efficiently navigating through the deluge of review information, a task that becomes increasingly complex due to the personalized nature of experiential products. Consumer demand for review information fluctuates based on their preferences for specific product features, a facet that existing research has yet to fully explore. This study seeks to bridge this gap by constructing a consumer class model and a class-based review helpfulness prediction model, leveraging the capabilities of LDA2Vec and LCR to achieve a personalized review ranking. The empirical results underscore the efficacy of the personalized ranking model in enhancing the perceived helpfulness of reviews. It significantly surpasses the helpfulness voting ranking mechanism in terms of review relevance, completeness, information diagnosticity, and consumer satisfaction. This research offers a robust methodology to identify the features of experiential products and to tailor the ranking of their reviews based on consumer preferences for these features. Additionally, it provides theoretical insights, modelling outcomes, and empirical evidence to alleviate the issue of review information overload through personalized review ranking.
%K 评论排序，体验型产品特征，LDA2Vec，LCR<br>Review Ranking
%K Experiential Product Features
%K LDA2Vec
%K LCR
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=79143