Abstract

Artificial intelligence (AI) and big data have driven e-commerce toward personalized recommendation systems, where data sharing is critical for performance optimization. However, existing mechanisms rely on static agreements and single incentives, leading to privacy risks, misaligned interests, and conflicts between platform data needs and user privacy protection. These challenges often result in inefficient, non-cooperative equilibrium that hinder sustainable ecosystem development. This study proposes an incentive-compatible dynamic data-sharing mechanism by modeling platform–user interaction as a multi-stage dynamic game. Core variables and payment functions are defined, and a sub-game perfect Nash equilibrium is derived to ensure stable cooperation. The framework integrates three modules—dynamic incentives, privacy protection, and adaptive feedback—forming a closed-loop system that responds to evolving AI-driven data requirements. Simulation experiments in Python and case validation using Amazon and Taobao demonstrate that the proposed mechanism improves key performance indicators by over 27%, with maximum gains reaching 107.89% compared to traditional approaches. The findings offer a practical pathway to balance data efficiency and privacy, supporting the development of a sustainable AI-enabled e-commerce ecosystem.

Download

Share

COinS