Syntax-BERT: Improving Pre-trained Transformers with Syntax Trees

Jiangang Bai, Yujing Wang, Yiren Chen, Yaming Yang, Jing Bai, Jing Yu, Yunhai Tong


Abstract
Pre-trained language models like BERT achieve superior performances in various NLP tasks without explicit consideration of syntactic information. Meanwhile, syntactic information has been proved to be crucial for the success of NLP applications. However, how to incorporate the syntax trees effectively and efficiently into pre-trained Transformers is still unsettled. In this paper, we address this problem by proposing a novel framework named Syntax-BERT. This framework works in a plug-and-play mode and is applicable to an arbitrary pre-trained checkpoint based on Transformer architecture. Experiments on various datasets of natural language understanding verify the effectiveness of syntax trees and achieve consistent improvement over multiple pre-trained models, including BERT, RoBERTa, and T5.
Anthology ID:
2021.eacl-main.262
Volume:
Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume
Month:
April
Year:
2021
Address:
Online
Editors:
Paola Merlo, Jorg Tiedemann, Reut Tsarfaty
Venue:
EACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
3011–3020
Language:
URL:
https://aclanthology.org/2021.eacl-main.262
DOI:
10.18653/v1/2021.eacl-main.262
Bibkey:
Cite (ACL):
Jiangang Bai, Yujing Wang, Yiren Chen, Yaming Yang, Jing Bai, Jing Yu, and Yunhai Tong. 2021. Syntax-BERT: Improving Pre-trained Transformers with Syntax Trees. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, pages 3011–3020, Online. Association for Computational Linguistics.
Cite (Informal):
Syntax-BERT: Improving Pre-trained Transformers with Syntax Trees (Bai et al., EACL 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.eacl-main.262.pdf
Code
 nkh2235/SyntaxBERT
Data
GLUEMultiNLIQNLISNLISSTSST-2