GARDRec: Decision-Level Graph Grounding for Large Language Model Recommendation

2026-08-01

AI Summary

This work addresses the limitations of existing large language model (LLM)-based recommendation approaches, which treat knowledge graphs merely as prompt evidence and struggle to effectively integrate temporal preferences, collaborative signals, and attribute matching. To overcome this, the authors propose GARDRec, a novel framework that constructs semantic-structural fused item representations via graph propagation, extracts personalized graph context from temporally weighted interaction histories and first-order neighbors, and aligns a frozen LLM through continuous multimodal prompting. At the decision layer, GARDRec uniquely incorporates explicit interaction features, inter-candidate attention, and constrained generation likelihood, enabling deep integration of graph-structured information into LLM-based recommendation decisions for the first time. Extensive experiments demonstrate that GARDRec significantly outperforms state-of-the-art methods across three benchmarks, with ablation studies confirming the contribution of each core component.

Abstract

Large language models (LLMs) offer new opportunities for recommendation by interpreting item descriptions, user instructions, and external knowledge through natural-language prompts. However, existing graph-augmented LLM recommenders often use knowledge graphs mainly as prompt-level evidence, leaving ranking decisions weakly constrained by structured user-item relations. This is problematic for next-item recommendation, where the model must compare candidates under the same user context while preserving temporal preference, collaborative signals, and attribute matches. To address this issue, we propose \emph{GARDRec}, a Graph-grounded Adaptive Reasoning and Decision-aware Recommendation framework for LLM-based next-item ranking. GARDRec constructs semantic-structural item representations from textual node features and graph propagation, derives personalized graph contexts from temporally weighted histories and first-order neighborhoods, and aligns graph-derived representations with a frozen LLM through continuous multimodal prompts. Explicit interaction and matching features are injected through late-stage decision branches, while inter-candidate attention and restricted generative likelihood support final ranking. Experiments on three public benchmarks with multiple LLM backbones show that GARDRec generally improves candidate-ranking performance over representative baselines. Ablation and diagnostic analyses verify the contributions of graph projection, neighborhood retrieval, explicit decision features, ranking loss, and generative calibration.

Problem

• large language models
• recommendation
• knowledge graphs
• next-item recommendation
• ranking

Innovation

• Graph-grounded reasoning
• Decision-aware recommendation
• Large language models
• Multimodal prompting
• Next-item ranking