Adapting Meta Knowledge Graph Information for Multi-Hop Reasoning over Few-Shot Relations
This addresses the challenge of robust reasoning in knowledge graphs for applications like query answering, where many relations have limited data, representing an incremental improvement over existing methods.
The paper tackles the problem of multi-hop knowledge graph reasoning for few-shot relations, where existing methods perform poorly due to insufficient training triples, and proposes Meta-KGR, a meta-learning approach that adapts knowledge from high-frequency relations to improve performance, achieving state-of-the-art results on Freebase and NELL datasets.
Multi-hop knowledge graph (KG) reasoning is an effective and explainable method for predicting the target entity via reasoning paths in query answering (QA) task. Most previous methods assume that every relation in KGs has enough training triples, regardless of those few-shot relations which cannot provide sufficient triples for training robust reasoning models. In fact, the performance of existing multi-hop reasoning methods drops significantly on few-shot relations. In this paper, we propose a meta-based multi-hop reasoning method (Meta-KGR), which adopts meta-learning to learn effective meta parameters from high-frequency relations that could quickly adapt to few-shot relations. We evaluate Meta-KGR on two public datasets sampled from Freebase and NELL, and the experimental results show that Meta-KGR outperforms the current state-of-the-art methods in few-shot scenarios. Our code and datasets can be obtained from https://github.com/ THU-KEG/MetaKGR.