Augmenting Context Representation with Triggers Knowledge for Relation Extraction

En Li, Shumin Shi*, Zhikun Yang, He Yan Huang

*Corresponding author for this work

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

Abstract

Relation Extraction (RE) requires the model to classify the correct relation from a set of relation candidates given the corresponding sentence and two entities. Recent work mainly studies how to utilize more data or incorporate extra context information especially with Pre-trained Language Models (PLMs). However, these models still face with the challenges of avoiding being affected by irrelevant or misleading words. In this paper, we propose a novel model to help alleviate such deficiency. Specifically, our model automatically mines the triggers of the sentence iteratively with the sentence itself from the previous iteration, and augment the semantics of the context representation from BERT with both entity pair and triggers skillfully. We conduct extensive experiments to evaluate the proposed model and effectively obtain empirical improvement in TACRED.

Original languageEnglish
Title of host publicationIntelligent Information Processing XI - 12th IFIP TC 12 International Conference, IIP 2022, Proceedings
EditorsZhongzhi Shi, Jean-Daniel Zucker, Bo An
PublisherSpringer Science and Business Media Deutschland GmbH
Pages124-135
Number of pages12
ISBN (Print)9783031039478
DOIs
Publication statusPublished - 2022
Event12th IFIP TC 12 International Conference on Intelligent Information Processing, IIP 2022 - Qingdao, China
Duration: 27 May 202230 May 2022

Publication series

NameIFIP Advances in Information and Communication Technology
Volume643 IFIP
ISSN (Print)1868-4238
ISSN (Electronic)1868-422X

Conference

Conference12th IFIP TC 12 International Conference on Intelligent Information Processing, IIP 2022
Country/TerritoryChina
CityQingdao
Period27/05/2230/05/22

Keywords

  • Context aware
  • Knowledge augment
  • Relation extraction
  • Triggers representation

Fingerprint

Dive into the research topics of 'Augmenting Context Representation with Triggers Knowledge for Relation Extraction'. Together they form a unique fingerprint.

Cite this