演讲主题:Mining Consumer-Generated Product Reviews to Automate Market Structure Analyses

　　主 讲 人：王欣 加拿大西安大略大学毅伟商学院营销方向助理教授

　　主 持 人：杨治 管理学院副院长

　　活动时间：2015年11月9日 下午14：30 -16：00

　　活动地点：管院121室

　　主讲人简介：王欣博士是毅伟商学院(Ivey Business School)营销方向助理教授。他的研究主要集中在选择模型，认知心理学，行为经济学，贝叶斯分析，机器学习，用户评论的文本分析等方面。他的研究成果发表于 Marketing Science, Journal of Consumer Research, Journal of Mathematical Psychology等顶级期刊上。

　　活动摘要：The increasing popularity of social media has led to exponential growth in user-generated content (UGC), such as consumer-generated product reviews, that could be highly beneficial for market intelligence. Due to the large volume and unstructured nature of UGC, it creates both a unique opportunity and notable challenges for academics and practitioners. To advance market structure analyses, this study combines marketing theories and advances in computational linguistics to obtain deeper consumer and market insights from free-form product reviews. An innovative text analytics method integrates text mining and sentiment analysis techniques to automate market structure analyses. To demonstrate the validity of this method, the authors conduct an intensive empirical study and implement a prototype system to analyze tablet computer reviews from Amazon.com. The results show that the proposed method provides high validity compared with extant market structure analysis methods and can explain market competition using a multi-attribute approach. The method offers additional insights, such as consumer product usage situations, multilevel attribute hierarchy, and consumer sentiments toward product attributes.