Elamur Research⌕
☰
home / papers / 2026 / Single-Pass, Depth-Selective Reading for Multi-Aspect Sentiment Analysis

Single-Pass, Depth-Selective Reading for Multi-Aspect Sentiment Analysis

Paper · 2026 · Yan Xia, Zhuangzhuang Pan, Amirrudin Kamsin et al.

Summary

The paper tackles aspect-term sentiment analysis (ATSA) for sentences that mention several aspects. Typical systems either run the encoder again for every aspect, so cost grows linearly with the number of aspects, or they compress all layers into one fixed vector, which loses aspect-specific detail. The authors propose DABS, which treats Transformer depth as a resource that each aspect can query. A sentence is encoded only once. A module called DORA then builds a shared, depth-ordered substrate. It keeps local cues such as negation by applying small depthwise-separable convolutions to the final layer, and it runs a GRU over the top K layers, from shallower to deeper. A second module, ACBS, lets each aspect choose which tokens to attend to and which depth levels to rely on, using a gated fusion, without re-encoding the sentence. On four standard benchmarks, DABS reaches accuracy and Macro-F1 comparable to strong systems. With two or more aspects per sentence, it lowers end-to-end compute by as much as 60%. This makes the approach relevant for high-throughput, resource-limited sentiment pipelines.

Key points

  • Evaluated on Lap14, Rest14, Rest15 and Rest16. The full model's mean Macro-F1 scores are 81.56, 84.87, 74.06 and 84.38, respectively. Code is released at github.com/panzhzh/acl-dabs.
  • Compute savings reach up to 60% when sentences have M ≥ 2 aspects. For single-aspect sentences, building the substrate adds a fixed overhead.
  • In the ablations, ACBS is the most important module. Removing token selection lowers Macro-F1 by 2.77 pp on average, and removing gated fusion lowers it by 2.21 pp.
  • Removing DepthGRU (K=6) lowers Macro-F1 on stress-test splits, with the largest effects on Multi-Aspect Conflict (76.88 → 71.32) and Complex Negation (80.62 → 77.69).
  • Depth-masking analyses show that adaptive depth selection helps most on harder linguistic cases such as negation and contrast. The method assumes gold aspect spans, so implicit or jointly extracted aspects are not covered.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓VinUniversityHanoi · Vietnam13 works · 56 citations🎓University of MalayaKuala Lumpur · Malaysia3 works · 273 citations

Also on the byline: Suzhou University of Technology.

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 works

🧭 Topics 1Research subjects and languages the work is about.Sentiment & emotion · 1,035 works

Cite

Cite as Yan Xia, Zhuangzhuang Pan, Amirrudin Kamsin and Chee Seng Chan. 2026. Single-Pass, Depth-Selective Reading for Multi-Aspect Sentiment Analysis. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.acl-long.667

@inproceedings{xia2026single,
  title = {{Single-Pass, Depth-Selective Reading for Multi-Aspect Sentiment Analysis}},
  author = {Yan Xia, Zhuangzhuang Pan, Amirrudin Kamsin and Chee Seng Chan},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.667},
  url = {https://doi.org/10.18653/v1/2026.acl-long.667}
}

References

45 references, 7 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Sangmin Bae, Jongwoo Ko, Hwanjun Song, and Se-Young Yun. 2023. Fast and robust early-exiting framework for autoregressive language models with synchronized parallel decoding. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 5910–5924, Singapore. Associ… find
  2. Xiaoyi Bao, Xiaotong Jiang, Zhongqing Wang, Yue Zhang, and Guodong Zhou. 2023a. Opinion tree parsing for aspect-based sentiment analysis. In Findings of the Association for Computational Linguistics: ACL 2023, pages 7971–7984, Toronto, Canada. Association for Computational Linguistics. find
  3. Xiaoyi Bao, Zhongqing Wang, and Guodong Zhou. 2023b. Exploring graph pre-training for aspect-based sentiment analysis. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 3623–3634, Singapore. Association for Computational Linguistics. find
  4. Aaditya Bodke, Avinoor Singh Kohli, Hemant Subhash Pardeshi, and Prathamesh Bhosale. 2025. PASTEL : Polarity-aware sentiment triplet extraction with LLM-as-a-judge. In Findings of the Association for Computational Linguistics: ACL 2025, pages 25523–25533, Vienna, Austria. Association for Computation… find
  5. Laura Cabello and Uchenna Akujuobi. 2024. It is simple sometimes: A study on improving aspect-based sentiment analysis performance. In Findings of the Association for Computational Linguistics: ACL 2024, pages 6597–6610, Bangkok, Thailand. Association for Computational Linguistics. find
  6. Heyan Chai, Ziyi Yao, Siyu Tang, Ye Wang, Liqiang Nie, Binxing Fang, and Qing Liao. 2023. Aspect-to-scope oriented multi-view contrastive learning for aspect-based sentiment analysis. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 10902–10913, Singapore. Association… find
  7. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. BERT: pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies,…
  8. Xuanwen Ding, Jie Zhou, Liang Dou, Qin Chen, Yuanbin Wu, Arlene Chen, and Liang He. 2024. Boosting large language models with continual learning for aspect-based sentiment analysis. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 4367–4377, Miami, Florida, USA. Associ… find
  9. Mostafa Elhoushi, Akshat Shrivastava, Diana Liskovich, Basil Hosmer, Bram Wasti, Liangzhen Lai, Anas Mahmoud, Bilge Acun, Saurabh Agarwal, Ahmed Roman, Ahmed Aly, Beidi Chen, and Carole-Jean Wu. 2024. LayerSkip: Enabling early exit inference and self-speculative decoding. In Proceedings of the 62nd…
  10. Rui Fan, Shu Li, Tingting He, and Yu Liu. 2025. Aspect-based sentiment analysis with syntax-opinion-sentiment reasoning chain. In Proceedings of the 31st International Conference on Computational Linguistics, pages 3123–3137, Abu Dhabi, UAE. Association for Computational Linguistics. find
  11. Ao Feng, Jiazhi Cai, Zhengjie Gao, and Xiaojie Li. 2023. Aspect-level sentiment classification with fused local and global context. J. Big Data, 10(1):176. find
  12. Zhibin Gou, Qingyan Guo, and Yujiu Yang. 2023. MvP: Multi-view prompting improves aspect sentiment tuple prediction. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 4380–4397, Toronto, Canada. Association for Computational Lin… find
  13. Bo He, Ruoyu Zhao, and Dali Tang. 2025. Cabilstm-bert: Aspect-based sentiment analysis model based on deep implicit feature extraction. Knowl. Based Syst., 309:112782. find
  14. Pengcheng He, Jianfeng Gao, and Weizhu Chen. 2023. Debertav3: Improving deberta using electra-style pre-training with gradient-disentangled embedding sharing. In The Eleventh International Conference on Learning Representations, ICLR 2023, Kigali, Rwanda, May 1-5, 2023. OpenReview.net.
  15. Nils Constantin Hellwig, Jakob Fehle, and Christian Wolff. 2025. Exploring large language models for the generation of synthetic training samples for aspect-based sentiment analysis in low resource settings. Expert Syst. Appl., 261:125514. find
  16. ZhongQuan Jian, Jiajian Li, Qingqiang Wu, and Junfeng Yao. 2024. Retrieval contrastive learning for aspect-level sentiment classification. Inf. Process. Manag., 61(1):103539. find
  17. Song Jin, Qing He, Yuji Wang, Nisuo Du, and Wenjing Lei. 2025. Aspect-based sentiment analysis with semantic and syntactic enhanced multi-layer fusion model. Eng. Appl. Artif. Intell., 159:111654. find
  18. Hongtao Liu, Xin Li, Wanying Lu, Kefei Cheng, and Xueyan Liu. 2024. Graph augmentation networks based on dynamic sentiment knowledge and static external knowledge graphs for aspect-based sentiment analysis. Expert Syst. Appl., 251:123981. find
  19. Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, and Veselin Stoyanov. 2019. Roberta: A robustly optimized BERT pretraining approach. CoRR, abs/1907.11692.
  20. Haoran Lv, Junyi Liu, Henan Wang, Yaoming Wang, Jixiang Luo, and Yaxiao Liu. 2023. Efficient hybrid generation framework for aspect-based sentiment analysis. In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics, pages 1007–1018, Dubrovnik, Cr… find
  21. Fukun Ma, Xuming Hu, Aiwei Liu, Yawen Yang, Shuang Li, Philip S. Yu, and Lijie Wen. 2023. Amr-based network for aspect-based sentiment analysis. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), ACL 2023, Toronto, Canada, July 9-14, 2… find
  22. Rajdeep Mukherjee, Nithish Kannen, Saurabh Pandey, and Pawan Goyal. 2023. CONTRASTE: Supervised contrastive pre-training with aspect-based prompts for aspect sentiment triplet extraction. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 12065–12080, Singapore. Associat… find
  23. Jihong Ouyang, Zhiyao Yang, Silong Liang, Bing Wang, Yimeng Wang, and Ximing Li. 2024. Aspect-based sentiment analysis with explicit sentiment augmentations. In Thirty-Eighth AAAI Conference on Artificial Intelligence, AAAI 2024, Thirty-Sixth Conference on Innovative Applications of Artificial Intel… find
  24. Jackson Petty, Sjoerd Steenkiste, Ishita Dasgupta, Fei Sha, Dan Garrette, and Tal Linzen. 2024. The impact of depth on compositional generalization in transformer language models. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Hu… find
  25. Maria Pontiki, Dimitris Galanis, Haris Papageorgiou, Ion Androutsopoulos, Suresh Manandhar, Mohammad AL-Smadi, Mahmoud Al-Ayyoub, Yanyan Zhao, Bing Qin, Orphée De Clercq, Véronique Hoste, Marianna Apidianaki, Xavier Tannier, Natalia Loukachevitch, Evgeniy Kotelnikov, Nuria Bel, Salud María Jiménez-Z… — SemEval-2016 Task 5: Aspect Based Sentiment Analysis
  26. Maria Pontiki, Dimitris Galanis, Haris Papageorgiou, Suresh Manandhar, and Ion Androutsopoulos. 2015. Semeval-2015 task 12: Aspect based sentiment analysis. In Proceedings of the 9th International Workshop on Semantic Evaluation, SemEval@NAACL-HLT 2015, Denver, Colorado, USA, June 4-5, 2015, pages 4…
  27. Maria Pontiki, Dimitris Galanis, John Pavlopoulos, Harris Papageorgiou, Ion Androutsopoulos, and Suresh Manandhar. 2014. Semeval-2014 task 4: Aspect based sentiment analysis. In Proceedings of the 8th International Workshop on Semantic Evaluation, SemEval@COLING 2014, Dublin, Ireland, August 23-24,…
  28. Zhiwen Ruan, Yixia Li, He Zhu, Longyue Wang, Weihua Luo, Kaifu Zhang, Yun Chen, and Guanhua Chen. 2025. LayAlign: Enhancing multilingual reasoning in large language models via layer-wise adaptive fusion and alignment strategy. In Findings of the Association for Computational Linguistics: NAACL 2025,… find
  29. Kevin Scaria, Himanshu Gupta, Siddharth Goyal, Saurabh Sawant, Swaroop Mishra, and Chitta Baral. 2024. InstructABSA: Instruction learning for aspect based sentiment analysis. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human L… find
  30. Yongsik Seo, Sungwon Song, Ryang Heo, Jieyong Kim, and Dongha Lee. 2024. Make compound sentences simple to analyze: Learning to split sentences for aspect-based sentiment analysis. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 11171–11184, Miami, Florida, USA. Assoc… find
  31. Chuming Shen, Wei Wei, Dong Wang, and Zhong-Hao Wang. 2025. Zero-shot cross-domain aspect-based sentiment analysis via domain-contextualized chain-of-thought reasoning. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 4558–4573, Suzhou, China. Association for Computati… find
  32. Xuefeng Shi, Min Hu, Fuji Ren, Piao Shi, and Satoshi Nakagawa. 2024. Aspect based sentiment analysis with instruction tuning and external knowledge enhanced dependency graph. Appl. Intell., 54(8):6415– 6432. find
  33. Paul F. Simmering and Paavo Huoviala. 2023. Large language models for aspect-based sentiment analysis. CoRR, abs/2310.18025. — [arXiv:2310.18025](https://arxiv.org/abs/2310.18025) find
  34. Xin Sun, Yongqing Mi, and Hongao Li. 2025. Enhancing aspect sentiment classification with dual-channel graph convolutional network. ACM Trans. Intell. Syst. Technol., 16(3). find
  35. Joachim Wagner and Jennifer Foster. 2023. Investigating the saliency of sentiment expressions in aspect-based sentiment analysis. In Findings of the Association for Computational Linguistics: ACL 2023, pages 12751–12769, Toronto, Canada. Association for Computational Linguistics. find
  36. Yanyan Wang, Qun Chen, Murtadha H. M. Ahmed, Zhaoqiang Chen, Jing Su, Wei Pan, and Zhanhuai Li. 2023. Supervised gradual machine learning for aspect-term sentiment analysis. Trans. Assoc. Comput. Linguistics, 11:723–739. find
  37. Shuo Yin and Guoqiang Zhong. 2024. Textgt: A double-view graph transformer on text for aspect-based sentiment analysis. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 19404–19412. find
  38. Mao Zhang, Yongxin Zhu, Zhen Liu, Zhimin Bao, Yunfei Wu, Xing Sun, and Linli Xu. 2023a. Span-level aspect-based sentiment analysis via table filling. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 9273–9284, Toronto, Canada.… find
  39. Siyu Zhang, Hongfang Gong, and Lina She. 2023b. An aspect sentiment classification model for graph attention networks incorporating syntactic, semantic, and knowledge. Knowl. Based Syst., 275:110662. find
  40. Guangmin Zheng, Jin Wang, Liang-Chih Yu, and Xuejie Zhang. 2024. Instruction tuning with retrieval-based examples ranking for aspect-based sentiment analysis. In Findings of the Association for Computational Linguistics: ACL 2024, pages 4777–4788, Bangkok, Thailand. Association for Computational Lin… find
  41. Yongqiang Zheng and Xia Li. 2024. You only read once: Constituency-oriented relational graph convolutional network for multi-aspect multi-sentiment classification. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 19715–19723. find
  42. Qihuang Zhong, Liang Ding, Juhua Liu, Bo Du, Hua Jin, and Dacheng Tao. 2023. Knowledge graph augmented network towards multiview representation learning for aspect-based sentiment analysis. IEEE Trans. Knowl. Data Eng., 35(10):10098–10111. find
  43. Linan Zhu, Xiangfan Chen, Xiaolei Guo, Chenwei Zhang, Zhechao Zhu, Zehai Zhou, and Xiangjie Kong. 2024. Pinpointing diffusion grid noise to enhance aspect sentiment quad prediction. In Findings of the Association for Computational Linguistics: ACL 2024, pages 3717–3726, Bangkok, Thailand. Associatio… find
  44. Zhihao Wang, Bo Zhang, Ru Yang, Chang Guo, and Maozhen Li. 2024. DAGCN: Distance-based and aspect-oriented graph convolutional network for aspect-based sentiment analysis. In Findings of the Association for Computational Linguistics: NAACL 2024, pages 1863–1876, Mexico City, Mexico. Association for… find
  45. Xiancai Xu, Jia-Dong Zhang, Lei Xiong, and Zhishang Liu. 2024. iACOS: Advancing implicit sentiment extraction with informative and adaptive negative examples. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technolo… find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/xia-2026-single-pass-depth-selective-reading/