Elamur Research⌕
☰
home / papers / 2025 / Diversity-oriented Data Augmentation with Large Language Models

Diversity-oriented Data Augmentation with Large Language Models

Paper · 2025 · Zaitian Wang, Jinghan Zhang, Xinhao Zhang et al.

Summary

Most text data augmentation methods add more training examples but do not check whether the new examples make the dataset more varied. This can produce redundant or label-corrupting samples that encourage overfitting. The authors propose DoAug, a framework built around a small LLM (Llama-3.2-1B-Instruct) that works as a paraphraser.

The model is trained in two steps:

  • Supervised fine-tuning on a paraphrase dataset, so it learns to rewrite sentences without changing their meaning.
  • DPO training on a preference dataset that favours varied rewrites over near-copies.

To keep LLM costs down, only an informative coreset of the target data is paraphrased. The paraphrases are then merged with the retained original samples.

Across 12 real-world text datasets, DoAug improved six diversity measures while keeping the augmented data close to the originals and their labels consistent. Downstream models improved by 10.52% on average, more than three points above the next-best baseline. The work matters because it treats diversity as an explicit goal of augmentation, alongside dataset size and label coherence.

Key points

  • Pipeline: SFT paraphraser, then DPO diversity tuning, then coreset selection, then merging paraphrases with the retained original data. Code is at github.com/CNICDS/DoAug.
  • Main result: an average downstream performance gain of 10.52% over 12 datasets, beating the runner-up baseline by more than three percentage points.
  • Robustness checks:
    • Replacing Llama-3.2-1B-Instruct with Qwen2.5-1.5B-Instruct kept diversity gains and gave comparable accuracy.
    • Using GPT-2 and T5-large instead of BERT as downstream classifiers on CoLA, MNLI and RCT also benefited from DoAug.
  • Evaluation of generated data: both a task-aware LLM and human judges were used to check correctness.
  • Limitations:
    • Evaluation is mostly English sentence classification, with two QA reasoning tasks and only one multilingual dataset.
    • Diversity metrics are not standardised, and human-centred diversity evaluation was not used.
    • Only two settings each were tried for temperature, prompts and LLM choice.
    • Possible bias and hallucination risks come from the LLM.

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

🏛️Chinese Academy of SciencesBeijing · China194 works · 23,906 citations🎓University of Chinese Academy of SciencesBeijing · China147 works · 8,480 citations🎓Portland State UniversityPortland · United States6 works · 53 citations🔬Computer Network Information CenterBeijing · China3 works · 22 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

Cite

Cite as Zaitian Wang, Jinghan Zhang, Xinhao Zhang, Kunpeng Liu, Pengfei Wang and Yuanchun Zhou. 2025. Diversity-oriented Data Augmentation with Large Language Models. Proceedings of the 63rd 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/v1/2025.acl-long.1084

@inproceedings{wang2025diversity,
  title = {{Diversity-oriented Data Augmentation with Large Language Models}},
  author = {Zaitian Wang, Jinghan Zhang, Xinhao Zhang, Kunpeng Liu, Pengfei Wang and Yuanchun Zhou},
  booktitle = {Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2025},
  doi = {10.18653/v1/2025.acl-long.1084},
  url = {https://doi.org/10.18653/v1/2025.acl-long.1084}
}

References

58 references, 24 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Ateret Anaby-Tavor, Boaz Carmeli, Esther Goldbraich, Amir Kantor, George Kour, Segev Shlomov, Naama Tepper, and Naama Zwerdling. 2020. Do not have enough data? deep learning to the rescue! In Proceedings of the AAAI Conference on Artificial Intelligence, volume 34, pages 7383–7390. find
  2. Francesco Barbieri, Luis Espinosa Anke, and Jose Camacho-Collados. 2022. Xlm-t: Multilingual language models in twitter for sentiment analysis and beyond. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 258– 266.
  3. Yonatan Belinkov and Yonatan Bisk. 2018. Synthetic and natural noise both break neural machine translation. In International Conference on Learning Representations. find
  4. Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, et al. 2020. Language models are few-shot learners. Advances in neural information processing systems, 33:1877–1901.
  5. Xunxin Cai, Meng Xiao, Zhiyuan Ning, and Yuanchun Zhou. 2023. Resolving the imbalance issue in hierarchical disciplinary topic inference via llm-based data augmentation. In 2023 IEEE International Conference on Data Mining (ICDM), pages 956–961. IEEE Computer Society. find
  6. Jan Cegin, Branislav Pecher, Jakub Simko, Ivan Srba, Mária Bieliková, and Peter Brusilovsky. 2024. Effects of diversity incentives on sample diversity and downstream model performance in llm-based text augmentation. In Proceedings of the 62nd Annual Meeting of the Association for Computational Lingu… find
  7. Jan Cegin, Jakub Simko, and Peter Brusilovsky. 2023. Chatgpt to replace crowdsourcing of paraphrases for intent classification: Higher diversity and comparable model robustness. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 1889–1905. find
  8. Michael Chen, Mike D’Arcy, Alisa Liu, Jared Fernandez, and Doug Downey. 2019. CODAH: An adversarially-authored question answering dataset for common sense. In Proceedings of the 3rd Workshop on Evaluating Vector Space Representations for NLP, pages 63–69, Minneapolis, USA. Association for Computatio…
  9. Xueqing Chen, Yang Gao, Ludi Wang, Wenjuan Cui, Jiamin Huang, Yi Du, and Bin Wang. 2024. Large language model enhanced corpus of co2 reduction electrocatalysts and synthesis procedures. Scientific Data, 11(1):347. find
  10. C Coleman, C Yeh, S Mussmann, B Mirzasoleiman, P Bailis, P Liang, J Leskovec, and M Zaharia. 2020. Selection via proxy: Efficient data selection for deep learning. In International Conference on Learning Representations (ICLR). find
  11. Varun Kumar, Ashutosh Choudhary, and Eunah Cho. 2020. Data augmentation using pre-trained transformer models. In Proceedings of the 2nd Workshop on Life-long Learning for Spoken Language Systems, pages 18–26.
  12. Haixing Dai, Zhengliang Liu, Wenxiong Liao, Xiaoke Huang, Yihan Cao, Zihao Wu, Lin Zhao, Shaochen Xu, Fang Zeng, Wei Liu, et al. 2025. Auggpt: Leveraging chatgpt for text data augmentation. IEEE Transactions on Big Data.
  13. Yi-An Lai, Xuan Zhu, Yi Zhang, and Mona Diab. 2020. Diversity, density, and homogeneity: Quantitative characteristic metrics for text collections. In Proceedings of the Twelfth Language Resources and Evaluation Conference, pages 1739–1746. find
  14. Franck Dernoncourt and Ji Young Lee. 2017. Pubmed 200k rct: a dataset for sequential sentence classification in medical abstracts. In Proceedings of the Eighth International Joint Conference on Natural Language Processing (Volume 2: Short Papers), pages 308–313.
  15. Yichuan Li, Kaize Ding, Jianling Wang, and Kyumin Lee. 2024. Empowering large language models for textual data augmentation. In Findings of the Association for Computational Linguistics: ACL 2024, pages 12734–12751, Bangkok, Thailand. Association for Computational Linguistics. find
  16. Bosheng Ding, Chengwei Qin, Ruochen Zhao, Tianze Luo, Xinze Li, Guizhen Chen, Wenhan Xia, Junjie Hu, Luu Anh Tuan, and Shafiq Joty. 2024. Data augmentation using llms: Data perspectives, learning paradigms and challenges. In Findings of the Association for Computational Linguistics ACL 2024, pages 1… find
  17. Zirui Liu, Haifeng Jin, Ting-Hsiang Wang, Kaixiong Zhou, and Xia Hu. 2021. Divaug: Plug-in automated data augmentation with explicit diversity maximization. In Proceedings of the IEEE/CVF international conference on computer vision, pages 4762–4770. find
  18. Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, et al. 2024. The llama 3 herd of models. arXiv preprint arXiv:2407.21783.
  19. Kenneth Enevoldsen, Isaac Chung, Imene Kerboua, Márton Kardos, Ashwin Mathur, David Stap, Jay Gala, Wissam Siblini, Dominik Krzemiński, Genta Indra Winata, et al. 2025. Mmteb: Massive multilingual text embedding benchmark. arXiv preprint arXiv:2502.13595.
  20. Raphael Gontijo-Lopes, Sylvia J Smullin, Ekin D Cubuk, and Ethan Dyer. 2020. Affinity and diversity: Quantifying mechanisms of data augmentation. arXiv preprint arXiv:2002.08973. — [arXiv:2002.08973](https://arxiv.org/abs/2002.08973) find
  21. Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2022. Lora: Low-rank adaptation of large language models. In International Conference on Learning Representations.
  22. Akbar Karimi, Leonardo Rossi, and Andrea Prati. 2021. Aeda: An easier data augmentation technique for text classification. Findings of the Association for Computational Linguistics: EMNLP 2021. find
  23. Jacob Devlin Ming-Wei Chang Kenton and Lee Kristina Toutanova. 2019. Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of NAACL-HLT, pages 4171–4186.
  24. Jens Kringelum, Sonny Kim Kjaerulff, Søren Brunak, Ole Lund, Tudor I Oprea, and Olivier Taboureau. 2016. Chemprot-3.0: a global chemical biology diseases mapping. Database, 2016:bav123. find
  25. Nikolaos Malandrakis, Minmin Shen, Anuj Goyal, Shuyang Gao, Abhishek Sethi, and Angeliki Metallinou. 2019. Controlled text generation for data augmentation in intelligent artificial agents. In Proceedings of the 3rd Workshop on Neural Generation and Translation, pages 90–98. find
  26. Junghyun Min, R Thomas McCoy, Dipanjan Das, Emily Pitler, and Tal Linzen. 2020. Syntactic data augmentation increases robustness to inference heuristics. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 2339–2352.
  27. Niklas Muennighoff, Nouamane Tazi, Loic Magne, and Nils Reimers. 2023. Mteb: Massive text embedding benchmark. In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics, pages 2014–2037.
  28. Yixin Nie, Adina Williams, Emily Dinan, Mohit Bansal, Jason Weston, and Douwe Kiela. 2020. Adversarial nli: A new benchmark for natural language understanding. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 4885–4901.
  29. Zhiyuan Ning, Chunlin Tian, Meng Xiao, Wei Fan, Pengyang Wang, Li Li, Pengfei Wang, and Yuanchun Zhou. 2024. Fedgcs: A generative framework for efficient client selection in federated learning via gradient-based optimization. In IJCAI. find
  30. Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. 2022. Training language models to follow instructions with human feedback. Advances in neural information processing systems, 35:27730–27744.
  31. Bo Pang and Lillian Lee. 2004. A sentimental education: Sentiment analysis using subjectivity summarization based on minimum cuts. In Proceedings of the 42nd Annual Meeting of the Association for Computational Linguistics (ACL-04), pages 271–278.
  32. Guy Tevet and Jonathan Berant. 2021. Evaluating the evaluation of diversity in natural language generation. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, pages 326–346.
  33. Mansheej Paul, Surya Ganguli, and Gintare Karolina Dziugaite. 2021. Deep learning on a data diet: Finding important examples early in training. Advances in neural information processing systems, 34:20596– 20607. find
  34. Jörg Tiedemann and Santhosh Thottingal. 2020. OPUS-MT — Building open translation services for the World. In Proceedings of the 22nd Annual Conferenec of the European Association for Machine Translation (EAMT), Lisbon, Portugal. find
  35. Geoff Pleiss, Tianyi Zhang, Ethan Elenberg, and Kilian Q Weinberger. 2020. Identifying mislabeled data using the area under the margin ranking. Advances in Neural Information Processing Systems, 33:17044– 17056. find
  36. Alex Wang, Amanpreet Singh, Julian Michael, Felix Hill, Omer Levy, and Samuel R Bowman. 2019. Glue: A multi-task benchmark and analysis platform for natural language understanding. In 7th International Conference on Learning Representations, ICLR 2019.
  37. Alec Radford, Karthik Narasimhan, Tim Salimans, Ilya Sutskever, et al. 2018. Improving language understanding by generative pre-training. OpenAI blog. find
  38. Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, Ilya Sutskever, et al. 2019. Language models are unsupervised multitask learners. OpenAI blog, 1(8):9. find
  39. Rafael Rafailov, Archit Sharma, Eric Mitchell, Christopher D Manning, Stefano Ermon, and Chelsea Finn. 2024. Direct preference optimization: Your language model is secretly a reward model. Advances in Neural Information Processing Systems, 36.
  40. Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, and Peter J Liu. 2020. Exploring the limits of transfer learning with a unified text-to-text transformer. Journal of machine learning research, 21(140):1–67.
  41. Rico Sennrich, Barry Haddow, and Alexandra Birch. 2016. Improving neural machine translation models with monolingual data. In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 86–96. find
  42. Swabha Swayamdipta, Roy Schwartz, Nicholas Lourie, Yizhong Wang, Hannaneh Hajishirzi, Noah A Smith, and Yejin Choi. 2020. Dataset cartography: Mapping and diagnosing datasets with training dynamics. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pa… find
  43. Alon Talmor, Jonathan Herzig, Nicholas Lourie, and Jonathan Berant. 2019. CommonsenseQA: A question answering challenge targeting commonsense knowledge. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies,…
  44. Qwen Team. 2024. Qwen2.5: A party of foundation models. find
  45. Ke Wang, Jiahui Zhu, Minjie Ren, Zeming Liu, Shiwei Li, Zongye Zhang, Chenkai Zhang, Xiaoyu Wu, Qiqi Zhan, Qingjie Liu, et al. 2024a. A survey on data synthesis and augmentation for large language models. arXiv preprint arXiv:2410.12896. — [arXiv:2410.12896](https://arxiv.org/abs/2410.12896) find
  46. Pengfei Wang, Wenhao Liu, Jiajia Wang, Yana Liu, Pengjiang Li, Ping Xu, Wentao Cui, Ran Zhang, Qingqing Long, Zhilong Hu, et al. 2025. sccom-pass: An integrated multi-species scrna-seq database for ai-ready. Advanced Science, page 2500870. find
  47. Zaitian Wang, Pengfei Wang, Kunpeng Liu, Pengyang Wang, Yanjie Fu, Chang-Tien Lu, Charu C. Aggarwal, Jian Pei, and Yuanchun Zhou. 2024b. A comprehensive survey on data augmentation. arXiv preprint arXiv:2405.09591. — [arXiv:2405.09591](https://arxiv.org/abs/2405.09591) find
  48. Jason Wei and Kai Zou. 2019. Eda: Easy data augmentation techniques for boosting performance on text classification tasks. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP…
  49. Janyce Wiebe, Theresa Wilson, and Claire Cardie. 2005. Annotating expressions of opinions and emotions in language. Language resources and evaluation, 39:165–210. find
  50. Ping Xu, Zhiyuan Ning, Pengjiang Li, Wenhao Liu, Pengyang Wang, Jiaxu Cui, Yuanchun Zhou, and Pengfei Wang. 2025. scsiameseclu: A siamese clustering framework for interpreting single-cell rna sequencing data. arXiv preprint arXiv:2505.12626. — [arXiv:2505.12626](https://arxiv.org/abs/2505.12626) find
  51. Kang Min Yoo, Dongju Park, Jaewook Kang, Sang-Woo Lee, and Woomyoung Park. 2021. Gpt3mix: Leveraging large-scale language models for text augmentation. In Findings of the Association for Computational Linguistics: EMNLP 2021, pages 2225–2239.
  52. Yu Yu, Shahram Khadivi, and Jia Xu. 2022. Can data diversity enhance learning generalization? In Proceedings of the 29th international conference on computational linguistics, pages 4933–4945. find
  53. Jinghan Zhang, Xiting Wang, Yiqiao Jin, Changyu Chen, Xinhao Zhang, and Kunpeng Liu. 2024a. Prototypical reward network for data-efficient rlhf. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 13871–13884. find
  54. Ran Zhang, Xuezhi Wang, Guannan Liu, Pengyang Wang, Yuanchun Zhou, and Pengfei Wang. 2025. Motif-oriented representation learning with topology refinement for drug-drug interaction prediction. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 1102–1110. find
  55. Xiang Zhang, Junbo Zhao, and Yann LeCun. 2015. Character-level convolutional networks for text classification. Advances in neural information processing systems, 28.
  56. Xinhao Zhang, Zaitian Wang, Lu Jiang, Wanfu Gao, Pengfei Wang, and Kunpeng Liu. 2024b. Tfwt: Tabular feature weighting with transformer. arXiv preprint arXiv:2405.08403. — [arXiv:2405.08403](https://arxiv.org/abs/2405.08403) find
  57. Haizhong Zheng, Rui Liu, Fan Lai, and Atul Prakash. 2023. Coverage-centric coreset selection for high pruning rates. In 11th International Conference on Learning Representations, ICLR 2023. find
  58. Yaowei Zheng, Richong Zhang, Junhao Zhang, Yanhan Ye, and Zheyan Luo. 2024. Llamafactory: Unified efficient fine-tuning of 100+ language models. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations), pages 400–410.

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wang-2025-diversity-oriented-data-augmentation-large/