什么样的评论更容易获得有用性投票——以亚马逊网站研究为例

发布时间：2018-09-03 11:21

【摘要】：【目的】购物网站评论系统中的投票机制有利于帮助消费者筛选出高质量评论。本文以评论有用性投票数为研究对象,探讨什么样的评论更容易获得有用性投票。【方法】以信息采纳理论和负面偏差理论为基础,基于亚马逊购物网站中的12 393条手机评论数据,结合文本分析与零膨胀负二项回归分析方法,从评论者信度、评论信息质量、评论极性三个方面探究评论有用性投票影响因素。【结果】研究结果表明,评论者有用性、评论信息量、评论回复数、极端评分、评论文本消极倾向对评论有用性投票数具有积极正向影响。评论者发表评论数、评论者是否确认购买对评论有用性投票数有负向影响。【局限】仅以手机这一搜索型产品为研究对象,研究结果欠缺普适性。【结论】本文研究成果对于改善电子商务评论排序系统具有借鉴意义。
[Abstract]:Objective: the voting mechanism in the review system of shopping websites helps consumers to screen out high-quality reviews. This paper takes the number of useful votes for comment as the research object, and discusses what kind of comment is easier to obtain the useful vote. [methods] based on the theory of information adoption and the theory of negative bias, Based on 12 393 mobile phone review data from Amazon shopping website, combining text analysis and zero inflation negative binomial regression analysis, this paper analyzes the reliability of the reviewer and the quality of the comment information. [results] the results showed that the reviewers' usefulness, the amount of comment information, the number of comment responses, the extreme score, The negative tendency of comment text has a positive effect on the number of useful votes. The number of comments made by the reviewer, whether the reviewer confirmed that the purchase had a negative impact on the number of votes on the usefulness of the review. [conclusion] the research results of this paper can be used for reference to improve the ranking system of electronic commerce reviews.
【作者单位】： 武汉大学信息管理学院;武汉大学电子商务研究与发展中心;
【基金】：国家自然科学基金项目“创新2.0超网络中知识流动和群集交互的协同研究”(项目编号:71373194)的研究成果之一
【分类号】：F713.36
，

本文编号：2219835