Adaptive Data Augmentation for Aspect Sentiment Quad Prediction (2401.06394v1)
Abstract: Aspect sentiment quad prediction (ASQP) aims to predict the quad sentiment elements for a given sentence, which is a critical task in the field of aspect-based sentiment analysis. However, the data imbalance issue has not received sufficient attention in ASQP task. In this paper, we divide the issue into two-folds, quad-pattern imbalance and aspect-category imbalance, and propose an Adaptive Data Augmentation (ADA) framework to tackle the imbalance issue. Specifically, a data augmentation process with a condition function adaptively enhances the tail quad patterns and aspect categories, alleviating the data imbalance in ASQP. Following previous studies, we also further explore the generative framework for extracting complete quads by introducing the category prior knowledge and syntax-guided decoding target. Experimental results demonstrate that data augmentation for imbalance in ASQP task can improve the performance, and the proposed ADA method is superior to naive data oversampling.
- “A survey on aspect-based sentiment analysis: Tasks, methods, and challenges,” TKDE, 2022.
- “Aspect sentiment quad prediction as paraphrase generation,” in EMNLP, 2021, pp. 9209–9219.
- “Dynamic multi-scale context aggregation for conversational aspect-based sentiment quadruple analysis,” arXiv preprint arXiv:2309.15476, 2023.
- “Exploring graph pre-training for aspect-based sentiment analysis,” in Findings-EMNLP, 2023, pp. 3623–3634.
- “Aspect-category-opinion-sentiment quadruple extraction with implicit aspects and opinions,” in ACL, 2021, pp. 340–350.
- Joseph Peper and Lu Wang, “Generative aspect-based sentiment analysis with contrastive learning and expressive structure,” in EMNLP, 2022, pp. 6089–6095.
- “LEGO-ABSA: A prompt-based task assemblable unified generative framework for multi-task aspect-based sentiment analysis,” in COLING, 2022, pp. 7002–7012.
- “Towards generative aspect-based sentiment analysis,” in EMNLP, 2021, pp. 504–510.
- “Easy-to-hard learning for information extraction,” in Findings-ACL, 2023, pp. 11913–11930.
- “MvP: Multi-view prompting improves aspect sentiment tuple prediction,” in ACL, 2023, pp. 4380–4397.
- “Aspect-based sentiment analysis with opinion tree generation,” in IJCAI, 2022, pp. 4044–4050.
- “Improving aspect sentiment quad prediction via template-order data augmentation,” in EMNLP, 2022, pp. 7889–7900.
- “Seq2Path: Generating sentiment tuples as paths of a tree,” in Findings-ACL, 2022, pp. 2215–2225.
- “Decoupling representation and classifier for long-tailed recognition,” in ICLR, 2020.
- “Learning to model the tail,” in NeurIPS, 2017.
- “COM-MRC: A COntext-masked machine reading comprehension framework for aspect sentiment triplet extraction,” in EMNLP, 2022, pp. 3230–3241.
- “Semantics-preserved data augmentation for aspect-based sentiment analysis,” in EMNLP, 2021, pp. 4417–4422.
- “Exploring the limits of weakly supervised pretraining,” in ECCV, 2018, p. 185–201.
- “Machine learning with oversampling and undersampling techniques: Overview study and experimental results,” in ICICS, 2020, pp. 243–248.
- “Target-aspect-sentiment joint detection for aspect-based sentiment analysis,” in AAAI, 2020, pp. 9122–9129.
- “Uncertainty-aware unlikelihood learning improves generative aspect sentiment quad prediction,” in Findings-ACL, 2023, pp. 13481–13494.
- “A unified one-step solution for aspect sentiment quad prediction,” in Findings-ACL, 2023, pp. 12249–12265.
- “SemEval-2015 task 12: Aspect based sentiment analysis,” in SemEval, 2015, pp. 486–495.
- “SemEval-2016 task 5: Aspect based sentiment analysis,” in SemEval, 2016, pp. 19–30.
- “Large-scale long-tailed recognition in an open world,” in CVPR, 2019.
Collections
Sign up for free to add this paper to one or more collections.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.