推荐算法设计和创作者激励
发布时间：2026-01-19
1月
20
|
时间和日期
|
2026-01-20 (星期二) 14:00 下午
-
15:00 下午
|
|
地点
|
综合教学楼D604会议室
|
| 标题 | 推荐算法设计和创作者激励 |
| 日期和时间 |
2026年1月20日（周二） 14:00 - 15:00 |
| 地点 | 综合教学楼D604会议室 |
| 主讲人 |
周舟 复旦大学 |
| 摘要 | Algorithmic feeds determine which creators receive traffic and which communities grow. We study a platform that chooses between an exclusive feed that protects an anchor creator’s followers and a redistributive feed that cross‑exposes those followers to entrants. The platform also sets revenue sharing, with a baseline share on public views and a possibly higher payout on follower views. Three results emerge. First, redistribution expands total exposure but lowers the anchor’s private return to effort; without targeted compensation on follower views, anchor effort—and sometimes participation—falls. Second, recommendation and revenue sharing are strategic complements: redistribution is sustainable precisely when follower traffic is paid at a higher rate, which restores the anchor’s incentive. Third, high, uniform sharing can backfire by tightening the cap on follower uplifts, blunting incentives where redistribution bites and reducing consumer surplus. Our model rationalizes the coexistence of “For You/Home” (redistributive) and “Following/Subscriptions” (exclusive), as well as how traffic allocation ties with revenue sharing mechanism. It yields testable predictions for when to protect follower traffic versus when to redistribute it with targeted incentives. |
| 主讲人简介 | 周舟，现任复旦大学管理学院信息管理与商业智能系副教授。他的研究领域包括数字平台的战略与设计、数字平台治理、数字平台和创新等方面。他的研究成果发表在MIS Quarterly、Academy of Management Review等国际知名期刊。 |
