Killing Two Birds with One Stone: Unifying Retrieval and Ranking with a Single Generative Recommendation Model

2025-04-23

AI Summary

Traditional multi-stage retrieval-then-ranking paradigms in recommender systems suffer from information fragmentation and loss across stages. To address this, we propose UniGRF—a unified generative retrieval-and-ranking framework that jointly models both stages as a single sequence-to-sequence generation task, enabling seamless information sharing. Our key contributions are: (1) a retrieval-ranking enhancement closed-loop mechanism that facilitates cross-stage collaborative refinement; (2) gradient-guided adaptive weighting optimization, achieving simultaneous performance gains with zero additional computational overhead; and (3) a model-agnostic design, fully compatible with existing generative recommendation architectures. Extensive experiments on multiple benchmark datasets demonstrate significant improvements over state-of-the-art methods. Ablation studies confirm the efficacy of cross-stage collaboration and strong scalability. UniGRF exhibits robust practicality and high potential for industrial deployment.

Technology Category

• Machine Learning: Learning Preferences or Rankings
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Natural Language Processing: Generation

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs

Abstract

In recommendation systems, the traditional multi-stage paradigm, which includes retrieval and ranking, often suffers from information loss between stages and diminishes performance. Recent advances in generative models, inspired by natural language processing, suggest the potential for unifying these stages to mitigate such loss. This paper presents the Unified Generative Recommendation Framework (UniGRF), a novel approach that integrates retrieval and ranking into a single generative model. By treating both stages as sequence generation tasks, UniGRF enables sufficient information sharing without additional computational costs, while remaining model-agnostic. To enhance inter-stage collaboration, UniGRF introduces a ranking-driven enhancer module that leverages the precision of the ranking stage to refine retrieval processes, creating an enhancement loop. Besides, a gradient-guided adaptive weighter is incorporated to dynamically balance the optimization of retrieval and ranking, ensuring synchronized performance improvements. Extensive experiments demonstrate that UniGRF significantly outperforms existing models on benchmark datasets, confirming its effectiveness in facilitating information transfer. Ablation studies and further experiments reveal that UniGRF not only promotes efficient collaboration between stages but also achieves synchronized optimization. UniGRF provides an effective, scalable, and compatible framework for generative recommendation systems.

Problem

• Unifies retrieval and ranking in recommendation systems
• Reduces information loss between stages
• Enhances inter-stage collaboration and optimization

Innovation

• Unified generative model for retrieval and ranking
• Ranking-driven enhancer for inter-stage collaboration
• Gradient-guided adaptive weighter for balanced optimization