Cross-Graph Fusion with Multi-Relational Syntactic and Semantic Branches for Multi-Aspect Sentiment Analysis
- 15 August 2025
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 473-477
Abstract
Aspect-based sentiment analysis aims to determine the sentiment polarity of specific aspects within a sentence. Existing methods often focus on a single aspect or a single type of relational information, limiting their ability to model complex dependencies. We propose a multi-branch cross-graph fusion network that uses a BERT encoder, followed by parallel multirelational syntactic and semantic branches. Each branch output is fused with BERT representations via cross-attention, and the resulting features are concatenated. Special encoding and masking strategies enable our model to classify the sentiment of all aspects simultaneously. Experiments on benchmark datasets confirm both its strong performance and interpretability.Keywords
This publication has 10 references indexed in Scilit:
- You Only Read Once: Constituency-Oriented Relational Graph Convolutional Network for Multi-Aspect Multi-Sentiment ClassificationProceedings of the AAAI Conference on Artificial Intelligence, 2024
- Relation construction for aspect-level sentiment classificationInformation Sciences, 2022
- Jointly Learning Aspect-Focused and Inter-Aspect Relations with Graph Convolutional Networks for Aspect Sentiment Analysis2020
- Relational Graph Attention Network for Aspect-based Sentiment Analysis2020
- SMART: Robust and Efficient Fine-Tuning for Pre-trained Natural Language Models through Principled Regularized Optimization2020
- RoBERTa: A Robustly Optimized BERT Pretraining Approach2019
- BERT: Pre-training of Deep Bidirectional Transformers for Language UnderstandingarXiv, 2019
- Learning Deep Features for Discriminative Localization2016
- Attention-based LSTM for Aspect-level Sentiment Classification2016
- SemEval-2014 Task 4: Aspect Based Sentiment Analysis2014