论文部分内容阅读
商品的网络在线评论及其相关的评论者信息都被认为是电子商务的重要资源,尽管大多数人都认为第三方评论网站是相对独立且可信度高的网站,大多数评论者也都是相对独立并不含偏见的,但是现有的研究也证实了在线评论中存在着不同类型的偏差。本文使用H-P滤波(HodrickPrescott filter)技术评估第三方网站评论中存在的偏差,通过ARMA模型和脉冲响应技术分析了评论偏差的特征,最后使用电影评论的面板数据,探讨了当期评论偏差大小的影响因素,证明了第三方评论顺序偏差的存在及其依赖因素。
While reviews of merchandise online and its associated commenter information are all considered important sources of e-commerce, most reviewers consider third-party review sites to be relatively independent and credible sites, and most commentators are Relative independence is not biased, but existing research also confirms the existence of different types of bias in online reviews. This paper uses HP filter (HodrickPrescottfilter) to evaluate the bias of third-party website comments, analyzes the characteristics of comment bias by ARMA model and impulse response technique, and finally discusses the influencing factors of the current comment bias using the panel data of movie reviews , Proving the existence of third-party comment sequence deviation and its dependence factors.