Hyper-KGGen: A Skill-Driven Knowledge Extractor for High-Quality Knowledge Hypergraph Generation

TL;DR

Hyper-KGGen introduces a skill-driven, dynamic framework for extracting high-quality knowledge hypergraphs from documents, effectively bridging the scenario gap and outperforming existing methods across diverse domains.

Contribution

It proposes a novel skill-evolving extraction process with a coarse-to-fine mechanism and an adaptive skill acquisition module, advancing hypergraph construction across multiple scenarios.

Findings

Hyper-KGGen outperforms baseline methods in hypergraph extraction tasks.

The adaptive skill library improves generalization across domains.

The HyperDocRED benchmark facilitates evaluation of hypergraph extraction methods.

Abstract

Knowledge hypergraphs surpass traditional binary knowledge graphs by encapsulating complex $n$-ary atomic facts, providing a more comprehensive paradigm for semantic representation. However, constructing high-quality hypergraphs remains challenging due to the \textit{scenario gap}: generic extractors struggle to generalize across diverse domains with specific jargon, while existing methods often fail to balance structural skeletons with fine-grained details. To bridge this gap, we propose \textbf{Hyper-KGGen}, a skill-driven framework that reformulates extraction as a dynamic skill-evolving process. First, Hyper-KGGen employs a \textit{coarse-to-fine} mechanism to systematically decompose documents, ensuring full-dimensional coverage from binary links to complex hyperedges. Crucially, it incorporates an \textit{adaptive skill acquisition} module that actively distills domain expertise into a Global Skill Library. This is achieved via a stability-based feedback loop, where extraction stability serves as a relative reward signal to induce high-quality skills from unstable traces and missed predictions. Additionally, we present \textbf{HyperDocRED}, a rigorously annotated benchmark for document-level knowledge hypergraph extraction. Experiments demonstrate that Hyper-KGGen significantly outperforms strong baselines, validating that evolved skills provide substantially richer guidance than static few-shot examples in multi-scenario settings.