Revisiting Fairness-aware Interactive Recommendation: Item Lifecycle as a Control Knob

TL;DR

This paper introduces a lifecycle-aware reinforcement learning framework for fairness in interactive recommendation systems, revealing a three-phase item lifecycle pattern and demonstrating improved fairness and engagement in real-world datasets.

Contribution

It uncovers a new three-phase item lifecycle pattern in short-video platforms and proposes LHRL, a hierarchical RL framework that balances fairness and accuracy using phase-specific dynamics.

Findings

LHRL significantly improves fairness and user engagement.

Lifecycle-aware rewards enhance existing RL models.

Item lifecycles follow a three-phase pattern, deviating from classical models.

Abstract

This paper revisits fairness-aware interactive recommendation (e.g., TikTok, KuaiShou) by introducing a novel control knob, i.e., the lifecycle of items. We make threefold contributions. First, we conduct a comprehensive empirical analysis and uncover that item lifecycles in short-video platforms follow a compressed three-phase pattern, i.e., rapid growth, transient stability, and sharp decay, which significantly deviates from the classical four-stage model (introduction, growth, maturity, decline). Second, we introduce LHRL, a lifecycle-aware hierarchical reinforcement learning framework that dynamically harmonizes fairness and accuracy by leveraging phase-specific exposure dynamics. LHRL consists of two key components: (1) PhaseFormer, a lightweight encoder combining STL decomposition and attention mechanisms for robust phase detection; (2) a two-level HRL agent, where the high-level policy imposes phase-aware fairness constraints, and the low-level policy optimizes immediate user engagement. This decoupled optimization allows for effective reconciliation between long-term equity and short-term utility. Third, experiments on multiple real-world interactive recommendation datasets demonstrate that LHRL significantly improves both fairness and user engagement. Furthermore, the integration of lifecycle-aware rewards into existing RL-based models consistently yields performance gains, highlighting the generalizability and practical value of our approach.