RADD: Retrieval-Augmented Discrete Diffusion for Multi-Modal Knowledge Graph Completion

TL;DR

RADD introduces a retrieval-augmented discrete diffusion framework for multi-modal knowledge graph completion, decoupling retrieval and reranking to improve accuracy and efficiency.

Contribution

The paper proposes a novel framework that separates retrieval and reranking in MMKGC, utilizing a relation-aware retriever and a discrete denoiser for better performance.

Findings

RADD achieves state-of-the-art results on three MMKGC benchmarks.

The decoupled retrieval and reranking approach improves recall and precision.

Ablation studies confirm the effectiveness of each component.

Abstract

Most multi-modal knowledge graph completion (MMKGC) models use one embedding scorer to do both retrieval over the full entity set and final decision making. We argue that this coupling is a core bottleneck: global high-recall search and local fine-grained disambiguation require different inductive biases. Therefore, we propose a Retrieval-Augmented Discrete Diffusion (RADD) framework to decouple retrieve and reranking for MMKGC. A relation-aware multimodal KGE retriever serves as both global retriever and distillation teacher, while a conditional discrete denoiser performs shortlist-level entity-identity generation for reranking. Training combines KGE supervision, denoising cross-entropy, and temperature-scaled distillation from the retriever to the denoiser. At inference, the designed Diff-Rerank first forms a top-$K$ shortlist with the retriever and then reranks it with the denoiser, ensuring that recall is a strict prerequisite for precision. Experiments on three MMKGC benchmarks show that RADD achieves the best performance and consistent gains over strong unimodal, multimodal, and LLM-based baselines, while ablations further verify the contribution of each component.