计量与数量经济学研讨会

题 目：Learning to Rank Under Evolving Consumer Reviews

主讲人：赵倞同 哥伦比亚大学博士生

时 间：2020年11月04日下午14:15-15:30

地 点：明德主楼734

摘 要：

In many electronic platforms, consumers leave reviews about their experiences. The body of reviews grows over time, providing a valuable source of information. Subsequent consumers as well as the platforms can make use of the information to learn about the quality of the products. The platforms, in particular, can use the information to better rank products, in order to meet consumer needs or to improve sales. We study how a platform can learn from consumer choices when they are presented with different product rankings, as well as from the reviews that they leave. Over time, the platform forms increasingly accurate estimates about product quality and consumer preferences. We provide an algorithm to simultaneously learn and provide personalized product rankings to consumers in order to maximize a cumulative expected reward. We use the notion of regret to quantify our algorithm’s performance.

主讲人简介： 赵倞同,哥伦比亚大学富的基金工程与应用科学学院运筹学和数据科学方向博士生。她的主要研究兴趣为具有丰富信息的动态环境下决策问题的优化方法，具体方向包括考虑消费者行为的服务系统，数据驱动的供应链管理，以及医疗运营和入院控制管理。目前, 她已经在国外知名学术期刊上发表论文3篇，并有多篇论文在国际顶级期刊的评审中。

bat365在线官网登录

计量与数量经济学系

2020年10月30日