Dependency-Type Weighted Graph Convolutional Network on End-to-End Aspect-Based Sentiment Analysis

Yusong Mu*, Shumin Shi

*Corresponding author for this work

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

1 Citation (Scopus)

Abstract

Previous studies consider little on using dependency-type messages in the E2E-ABSA task. Studies using dependency-type messages just contact the dependency-type message and word embedding vectors, which may not fully fuse the context feature and information from the dependency type. This paper proposes a new model called Dependency-Type Weighted Graph Convolution Network (DTW-GCN) to compose dependency-type messages and word embedding. We use a type-weighted matrix to combine the dependency-type message, and DTW-GCN could fuse the dependency-type message and word embedding vectors. Experiments conducted on three benchmark datasets verify the effectiveness of our model.

Original languageEnglish
Title of host publicationIntelligent Information Processing XII - 13th IFIP TC 12 International Conference, IIP 2024, Proceedings
EditorsZhongzhi Shi, Jim Torresen, Shengxiang Yang
PublisherSpringer Science and Business Media Deutschland GmbH
Pages46-57
Number of pages12
ISBN (Print)9783031579189
DOIs
Publication statusPublished - 2024
Event13th IFIP TC 12 International Conference on Intelligent Information Processing, IIP 2024 - Shenzhen, China
Duration: 3 May 20246 May 2024

Publication series

NameIFIP Advances in Information and Communication Technology
Volume704 IFIPAICT
ISSN (Print)1868-4238
ISSN (Electronic)1868-422X

Conference

Conference13th IFIP TC 12 International Conference on Intelligent Information Processing, IIP 2024
Country/TerritoryChina
CityShenzhen
Period3/05/246/05/24

Keywords

  • Dependency syntactic knowledge
  • End-to-End Aspect-Based Sentiment Analysis
  • Graph Convolutional Network

Fingerprint

Dive into the research topics of 'Dependency-Type Weighted Graph Convolutional Network on End-to-End Aspect-Based Sentiment Analysis'. Together they form a unique fingerprint.

Cite this