Aspect-Based Sentiment Analysis with Explicit Sentiment Augmentations

Jihong Ouyang,Zhiyao Yang,Silong Liang,Bing Wang,Yimeng Wang,Ximing Li

2023-12-18

Abstract:Aspect-based sentiment analysis (ABSA), a fine-grained sentiment classification task, has received much attention recently. Many works investigate sentiment information through opinion words, such as ''good'' and ''bad''. However, implicit sentiment widely exists in the ABSA dataset, which refers to the sentence containing no distinct opinion words but still expresses sentiment to the aspect term. To deal with implicit sentiment, this paper proposes an ABSA method that integrates explicit sentiment augmentations. And we propose an ABSA-specific augmentation method to create such augmentations. Specifically, we post-trains T5 by rule-based data. We employ Syntax Distance Weighting and Unlikelihood Contrastive Regularization in the training procedure to guide the model to generate an explicit sentiment. Meanwhile, we utilize the Constrained Beam Search to ensure the augmentation sentence contains the aspect terms. We test ABSA-ESA on two of the most popular benchmarks of ABSA. The results show that ABSA-ESA outperforms the SOTA baselines on implicit and explicit sentiment accuracy.

Computation and Language,Artificial Intelligence

What problem does this paper attempt to address?

The paper aims to address the issue of implicit sentiment analysis in the task of fine-grained sentiment classification. Specifically:

- **Research Background**: In Aspect-Based Sentiment Analysis (ABSA), many existing methods rely on explicit sentiment lexicons (such as "good" and "bad") to extract sentiment information. However, in real-world datasets, there is a large amount of implicit sentiment data that lacks obvious opinion words, making it difficult to determine sentiment polarity.

- **Proposed Method**: To handle the problem of implicit sentiment, the authors propose a new ABSA framework—ABSA-Explicit Sentiment Augmentations (ABSA-ESA), which provides more sentiment clues by generating augmented sentences containing explicit sentiment expressions. Specifically, they design an ABSA-specific sentiment generation method to create augmented data with clear sentiment expressions.

- **Technical Details**:
- The T5 model is used for post-training, and rule-based data selection is employed to ensure that the generated augmented sentences meet the requirements.
- Three strategies (syntactic distance weighting, dissimilar contrastive regularization, and constrained beam search) are introduced during training to guide the model in generating explicit opinion words with the same sentiment polarity as the input sentence and to ensure that the augmented data is associated with the target aspect.

- **Experimental Results**: Experiments on two benchmark datasets show that ABSA-ESA outperforms existing state-of-the-art methods in both explicit and implicit sentiment accuracy. Additionally, ablation studies validate the effectiveness of each component, and case studies demonstrate the specific effects of the augmented sentences.