Context-Aware Adapter Tuning for Few-Shot Relation Learning in Knowledge Graphs

TL;DR

This paper introduces RelAdapter, a context-aware adapter that improves few-shot relation learning in knowledge graphs by incorporating relation-specific adaptation and contextual information, leading to superior performance on benchmark datasets.

Contribution

The paper presents a novel, lightweight, context-aware adapter for meta-learning in knowledge graphs, addressing distribution mismatch and enhancing relation-specific adaptation.

Findings

RelAdapter outperforms state-of-the-art methods on three benchmark KGs.

Incorporating contextual information improves relation adaptation.

RelAdapter is parameter-efficient and effective in few-shot scenarios.

Abstract

Knowledge graphs (KGs) are instrumental in various real-world applications, yet they often suffer from incompleteness due to missing relations. To predict instances for novel relations with limited training examples, few-shot relation learning approaches have emerged, utilizing techniques such as meta-learning. However, the assumption is that novel relations in meta-testing and base relations in meta-training are independently and identically distributed, which may not hold in practice. To address the limitation, we propose RelAdapter, a context-aware adapter for few-shot relation learning in KGs designed to enhance the adaptation process in meta-learning. First, RelAdapter is equipped with a lightweight adapter module that facilitates relation-specific, tunable adaptation of meta-knowledge in a parameter-efficient manner. Second, RelAdapter is enriched with contextual information about the target relation, enabling enhanced adaptation to each distinct relation. Extensive experiments on three benchmark KGs validate the superiority of RelAdapter over state-of-the-art methods.