客户购买记录可以预测客户评论行为吗?  

Does a Customer's Purchase Behavior Can be used to Predict His Review Behavior?

作　　者：刘慧丽[1,2] 赵萌[1,2] 齐佳音[2,3] Liu Huili Zhao Meng Qi Jiayin(School of Economics and Management, Beijing University of Posts and Telecommunications, Beijing 100876 Key Laboratory of Trustworthy Distributed Computing and Service （BUFF）, Ministry of Education, Beijing 100876 Shanghai University of International Business and Economics, Shanghai 201620)

机构地区：[1]北京邮电大学经济管理学院,北京100876 [2]可信分布式计算与服务教育部重点实验室,北京邮电大学,北京100876 [3]上海对外经贸大学,上海201620

出　　处：《情报杂志》2017年第2期194-200,共7页Journal of Intelligence

基　　金：国家自然科学基金项目"半契约情景下的客户终生价值建模及营销决策研究"(编号:71171023);国家自然科学基金重点项目"面向不确定性的Web2.0用户创作内容管理研究"(编号:71231002)

摘　　要：[目的/意义]Web2.0环境下互联网中出现大量的用户创作内容(User generated content,UGC)。这一改变给传统的客户关系管理带来了深刻影响,越来越多的学者开始关注客户评论,但是以前的研究大部分关注评论对购买行为的影响,而忽略了购买对评论的反作用。因此,研究购买行为对评论行为的影响作用具有十分重要的意义。[方法/过程]基于大众点评网数据,综合考虑用户自身因素、当前评论行为、客户购买记录,研究了购买记录对评论行为的影响作用,并构建Logit回归分析模型,预测客户在下一阶段的评论行为。[结果/结论]研究结果表明购买记录可以预测客户的评论行为,并采用十重交叉验证了模型的稳定性,为Web2.0用户创作内容的研究提供了理论基础。[ Purpose/Significance ] With the development of Web 2.0, traditional customers have increasingly transferred to online pur- chase and created a large volume of User Generated Content （UGC） on the Internet, which had a great impact on the traditional customer relationship management and attracted plenty of scholars＂ attention to customer review. Most of the previous researches focus on the influ- ence of customer review on customer＇s purchase behavior, while there are very limited researches exploring the impact on the reverse direc- tion. Therefore, it is of great significance to study the effect of customer＇s purchase behavior on his review behavior. [ Method/Process] Based on data from Dianping. com, a famous comprehensive website which contains review and purchase platforms, a Logit regression model is built, taking into consideration of customer＇s own factors, review and purchase behavior and identifying the impact of user＇s own purchase behavior on its review behavior. And customer＇s review churn in next stage is predicted. [ Result/Condusion] The results show that customer＇s purchase behavior can predict his review behavior. The stability of the model proposed is proved via Ten-fold cross-vali- dation. The results can provide a theoretical basis for research on User Generated Content.

关 键 词：评论行为 购买行为 LOGIT回归模型 流失预测 十重交叉验证 

分 类 号：F270[经济管理—企业管理] C934[经济管理—国民经济]