Creator Incentives in Recommender Systems: A Cooperative Game-Theoretic Approach for Stable and Fair Collaboration in Multi-Agent Bandits

2026-04-09

AI Summary

This work addresses the misaligned incentives among multiple content creators in recommendation systems, where user feedback coupling complicates fair credit attribution. To tackle this issue, the paper introduces cooperative game theory into a multi-agent stochastic linear bandit framework for the first time, defining coalition value via negative cumulative regret and formulating a transferable utility cooperative game. Theoretical analysis shows that under homogeneous agents, the resulting game is convex and its Shapley value lies in the core. For heterogeneous settings, the authors propose a regret-based payoff allocation rule that satisfies core membership and aligns with most Shapley axioms. Experiments on the MovieLens-100k dataset demonstrate that this rule consistently achieves fairness, practicality, and stability across diverse algorithms and configurations.

Technology Category

• Game Theory and Economic Paradigms: Cooperative Game Theory
• Multiagent Systems: Mechanism Design
• Machine Learning: Online Learning & Bandits

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Economics and fairness of platforms and recommendation systems
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

User interactions in online recommendation platforms create interdependencies among content creators: feedback on one creator's content influences the system's learning and, in turn, the exposure of other creators'contents. To analyze incentives in such settings, we model collaboration as a multi-agent stochastic linear bandit problem with a transferable utility (TU) cooperative game formulation, where a coalition's value equals the negative sum of its members'cumulative regrets. We show that, for identical (homogenous) agents with fixed action sets, the induced TU game is convex under mild algorithmic conditions, implying a non-empty core that contains the Shapley value and ensures both stability and fairness. For heterogeneous agents, the game still admits a non-empty core, though convexity and Shapley value core-membership are no longer guaranteed. To address this, we propose a simple regret-based payout rule that satisfies three out of the four Shapley axioms and also lies in the core. Experiments on MovieLens-100k dataset illustrate when the empirical payout aligns with -- and diverges from -- the Shapley fairness across different settings and algorithms.

Problem

• recommender systems
• creator incentives
• multi-agent bandits
• cooperative game theory
• fairness

Innovation

• cooperative game theory
• multi-agent bandits
• Shapley value
• regret-based payout
• creator incentives