Knowledge-Enhanced Prototypical Network with Structural Semantics for Few-Shot Relation Classification

  • Yanhu Li
  • , Taolin Zhang
  • , Dongyang Li
  • , Xiaofeng He*
  • *Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Few-shot relation classification (RC) aims to determine the labeled relation between two entities in a given sentence using only a few training instances. Previous studies integrate models with explicit triple knowledge, using the inherent concepts of entities to improve the instance representation. However, these studies neglect the implicit structural knowledge present in the knowledge graph (KG). In this paper, we present SKProto, a knowledge-enhanced prototypical network that leverages deep structured semantic knowledge from the multi-hop neighbors of entity-linked concepts. Specifically, we propose a concept-guided hybrid attention mechanism to learn implicit structural semantic knowledge for enhancing the context-aware instance representation. To further distinguish subtle semantic differences among the concepts, the multi-granularity semantic distinction approach is proposed to construct the negative samples with various difficulties (i.e. hard, medium, and easy) based on the conceptual hierarchical structure. Experimental results on the FewRel 2.0 benchmark show that SKProto outperforms state-of-the-art models. We also demonstrate that SKProto has better robustness than other competitive models in low-shot scenarios.

Original languageEnglish
Title of host publicationAdvances in Knowledge Discovery and Data Mining - 27th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2023, Proceedings
EditorsHisashi Kashima, Tsuyoshi Ide, Wen-Chih Peng
PublisherSpringer Science and Business Media Deutschland GmbH
Pages138-149
Number of pages12
ISBN (Print)9783031333798
DOIs
StatePublished - 2023
Event27th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2023 - Osaka, Japan
Duration: 25 May 202328 May 2023

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume13937 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference27th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2023
Country/TerritoryJapan
CityOsaka
Period25/05/2328/05/23

Keywords

  • Contrastive Learning
  • Few-shot Learning
  • Knowledge Graph
  • Relation Classification

Fingerprint

Dive into the research topics of 'Knowledge-Enhanced Prototypical Network with Structural Semantics for Few-Shot Relation Classification'. Together they form a unique fingerprint.

Cite this