THOR: Inductive Link Prediction over Hyper-Relational Knowledge Graphs

2026-02-05

AI Summary

This work proposes THOR, the first fully inductive framework for link prediction in hyper-relational knowledge graphs, addressing the limitation of existing methods that are largely confined to transductive settings and fail to generalize to unseen entities and relations. THOR achieves this by constructing entity- and relation-agnostic base graphs—namely, an entity base graph and a relation base graph—and employs a parallel graph neural network encoder coupled with a Transformer decoder. A masked training strategy is introduced to preserve structural invariance while enhancing cross-graph generalization. Evaluated across 12 datasets, THOR substantially outperforms current approaches, yielding relative MRR improvements of 66.1%, 55.9%, and 20.4% over the best rule-based, semi-inductive, and fully inductive baselines, respectively.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion
• Knowledge Representation and Reasoning: Other Foundations of Knowledge Representation & Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Knowledge graphs (KGs) have become a key ingredient supporting a variety of applications. Beyond the traditional triplet representation of facts where a relation connects two entities, modern KGs observe an increasing number of hyper-relational facts, where an arbitrary number of qualifiers associated with a triplet provide auxiliary information to further describe the rich semantics of the triplet, which can effectively boost the reasoning performance in link prediction tasks. However, existing link prediction techniques over such hyper-relational KGs (HKGs) mostly focus on a transductive setting, where KG embedding models are learned from the specific vocabulary of a given KG and subsequently can only make predictions within the same vocabulary, limiting their generalizability to previously unseen vocabularies. Against this background, we propose THOR, an inducTive link prediction technique for Hyper-relational knOwledge gRaphs. Specifically, we first introduce both relation and entity foundation graphs, modeling their fundamental inter- and intra-fact interactions in HKGs, which are agnostic to any specific relations and entities. Afterward, THOR is designed to learn from the two foundation graphs with two parallel graph encoders followed by a transformer decoder, which supports efficient masked training and fully-inductive inference. We conduct a thorough evaluation of THOR in hyper-relational link prediction tasks on 12 datasets with different settings. Results show that THOR outperforms a sizable collection of baselines, yielding 66.1%, 55.9%, and 20.4% improvement over the best-performing rule-based, semi-inductive, and fully-inductive techniques, respectively. A series of ablation studies also reveals our key design factors capturing the structural invariance transferable across HKGs for inductive tasks.

Problem

• inductive link prediction
• hyper-relational knowledge graphs
• knowledge graph reasoning
• out-of-vocabulary generalization

Innovation

• inductive link prediction
• hyper-relational knowledge graphs
• foundation graphs
• graph neural networks
• transformer decoder