Research on aspect level text sentiment analysis method based on BERT

Jinbo Liang; Chao Chen; Jiong Liu

doi:10.1117/12.2659120

28 October 2022 Research on aspect level text sentiment analysis method based on BERT

Jinbo Liang, Chao Chen, Jiong Liu

Proceedings Volume 12453, Third International Conference on Computer Communication and Network Security (CCNS 2022); 124530T (2022) https://doi.org/10.1117/12.2659120
Event: Third International Conference on Computer Communication and Network Security (CCNS 2022), 2022, Hohhot, China

Abstract

The traditional word vector model cannot generate word vectors with polysemous features. The paper uses the BERT model to obtain word vectors, analyzes the modeling ability of the model in text sentiment analysis, and improves the Bert output structure. In order to obtain more text vector information, the vector output structure of BERT model is improved, the original single-layer output is replaced with multi-layers hidden state output, and the hidden information of these different hidden layers is fused and embedded into the downstream task to perform text analyze. The results show that the method proposed in this paper has a certain improvement in the text classification task.

Citation Download Citation

Jinbo Liang, Chao Chen, and Jiong Liu "Research on aspect level text sentiment analysis method based on BERT", Proc. SPIE 12453, Third International Conference on Computer Communication and Network Security (CCNS 2022), 124530T (28 October 2022); https://doi.org/10.1117/12.2659120

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available