RECOMMENDER SYSTEMS AS MECHANISMS FOR SOCIAL LEARNING

• 成果类型: Article
• 署名作者: Che, Yeon-Koo; Horner, Johannes
• 署名单位: Columbia University; Yale University; Universite de Toulouse; Universite Toulouse 1 Capitole; Toulouse School of Economics
• 刊物名称: QUARTERLY JOURNAL OF ECONOMICS
• ISSN/ISSBN: 0033-5533
• DOI: 10.1093/qje/qjx044
• 发表日期: 2018
• 页码: 871-925
• 关键词: Information disclosure; experimentation; bandits; MARKET

摘要

This article studies how a recommender system may incentivize users to learn about a product collaboratively. To improve the incentives for early exploration, the optimal design trades off fully transparent disclosure by selectively overrecommending the product (or spamming) to a fraction of users. Under the optimal scheme, the designer spams very little on a product immediately after its release but gradually increases its frequency; she stops it altogether when she becomes sufficiently pessimistic about the product. The recommender's product research and intrinsic/naive users seed incentives for user exploration and determine the speed and trajectory of social learning. Potential applications for various Internet recommendation platforms and implications for review/ratings inflation are discussed.