With the rapid development of e-commerce enterprises, herding phenomenon gradually revealed. Using the characteristics and indicators of mathematical statistics and adopting the data of an e-commerce site to assess the results of the best businesses , this paper attempts to quantitatively analyze and define the herding phenomenon in the e-commerce business operations. Consistency and homo-geneous continuity features show the members' characteristics participating in herding and non-Poisson distribution and sudden increase features show the formation process of herding. The results show that the membership level, gender, amount of consumer, number of fans, and number of writing topics will not cause a significant impact on the merchant's score. The number of reviews wrote with days, weeks, months as the time period disobeys Poisson distribution. At 20% of the time points before and after the formation process, the sudden increase feature appeared. Homogeneous continuity and similar continuity are very obvious.%随着电子商务企业的快速发展，羊群效应现象逐步显现。文章运用数理统计的特征和指标，从某电子商务网站对最佳商户的评定结果入手，定量分析和描述电子商务运营中的羊群效应现象，并以一致性特征、非泊松分布特征、突增性特征、同质连续性特征四个指标对此进行分析。其中一致性、同质连续性呈现出了羊群效应成员参与的群体特征；非泊松分布、突增性呈现出了羊群效应产生形成的过程特征。研究结果表明，会员级别、性别、消费额度、拥有粉丝数、撰写点评数、发表主题贴数均不会对商户的星级打分造成显著影响；以天、周、月为周期的前来点评的次数不服从Poisson分布；在观察期的20%时间点位前后，出现了人数突增情况；同质连续性占比和相近连续性占比两个指标均十分明显。
展开▼
