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

TL;DR

The paper tackles information loss and inefficiency in traditional two-stage recommender systems by unifying retrieval and ranking into a single autoregressive generative model, treating both tasks as sequence generation outputs distinguished by position. It introduces a ranking-driven enhancer to generate high-quality cross-stage samples and a gradient-guided adaptive weighter to synchronize optimization across stages, enabling end-to-end learning with shared parameters. Empirical results on MovieLens and Amazon-Books show UniGRF outperforms state-of-the-art baselines, with ablations confirming the contributions of both enhancer and weighter and analyses demonstrating scalable improvements with model size. The approach offers a scalable, model-agnostic pathway toward fully integrated generative recommendation, with potential extensions to include prerank and rerank stages in the future.

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.

Figures (4)

• Figure 1: Overview of UniGRF. This framework unifies retrieval and ranking into a generative recommendation model, with a ranking-driven enhancer and a gradient-guided adaptive weighter for efficient collaboration and synchronized optimization.
• Figure 2: Impact of varying the number of hard-to-detect disliked items $m$.
• Figure 3: Effect of synchronized stage optimization on unified generative framework performance.
• Figure 4: Effect of parameter expansion on model loss and recommendation performance.