Elamur ResearchβŒ•
☰
home / papers / 2026 / Structured Legal Document Generation in India: A Model-Agnostic Wrapper Approach with VidhikDastaavej

Structured Legal Document Generation in India: A Model-Agnostic Wrapper Approach with VidhikDastaavej

Summary

Drafting private legal documents such as contracts and agreements takes a lot of work. Legal NLP in India has paid it little attention, mainly because these texts are confidential and no public training data exists. The authors close this gap with VidhikDastaavej, a collection of anonymized private drafts. They obtained it through a partnership with an Indian law firm, and it spans 133 document categories. They also propose the Model-Agnostic Wrapper (MAW), which needs no task-specific fine-tuning. It works in two steps: it first plans a document's section titles from the user's instructions, then writes each section in turn with retrieval-based prompts. Because of this design, it can run on top of any open, commercial or privately hosted LLM. Outputs were scored with lexical, semantic and LLM-based metrics, and legal professionals also rated them, with agreement between raters checked. The authors report that MAW drafts are more factually accurate, coherent and complete than those from fine-tuned baselines. The work matters for two reasons: it is the first benchmark for Indian private legal drafting, and its method stays usable as new models keep appearing.

Key points

  • VidhikDastaavej contains 11,825 documents: 11,692 for training and 133 for testing, one per category across all 133 categories. Average length is about 5,799 words in train and 7,465 in test, and the longest document has 98,607 words.
  • Annotation was semi-automatic. Mixtral assigned initial categories and LLaMA-3.1 produced section headers, and practicing legal experts then checked both.
  • Three independent legal experts rated generated drafts on a 1–10 Likert scale for two criteria: factual accuracy (following instructions, no hallucinations) and completeness (coverage of required sections and details). Inter-annotator agreement was reported.
  • The authors also built an interactive Human-in-the-Loop (HITL) interface. Users choose a document type, adjust sections and generate a structured draft.
  • The dataset, models and code are released via the GitHub repository ShubhamKumarNigam/VidhikDastaavej.

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

Fields: Linguistics & languages Β· Law

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.63317/2wswvwyqxt4k

@inproceedings{nigam2026structured,
  title = {{Structured Legal Document Generation in India: A Model-Agnostic Wrapper Approach with VidhikDastaavej}},
  author = {Nigam, Shubham Kumar and Balaramamahanthi, Deepak Patnaik and Shallum, Noel and Ghosh, Kripabandhu and Bhattacharya, Arnab},
  booktitle = {Proceedings of the Fifteenth Language Resources and Evaluation Conference},
  year = {2026},
  doi = {10.63317/2wswvwyqxt4k},
  url = {https://doi.org/10.63317/2wswvwyqxt4k}
}

References

42 references, 3 of them with a card of ours. β†— leads to the work, Β«findΒ» searches for one we have no id for yet.

  1. Satanjeev Banerjee and Alon Lavie. 2005. ME-TEOR: An automatic metric for MT evaluation with improved correlation with human judgments. In Proceedings of the ACL Workshop on Intrinsic and Extrinsic Evaluation Measures for Machine Translation and/or Summarization, pages 65–72, Ann Arbor, Michigan. Association for Computational Linguistics. find
  2. Jacob Benesty, Jingdong Chen, Yiteng Huang, and Israel Cohen. 2009. Pearson correlation coefficient. In Noise reduction in speech processing, pages 1–4. Springer. β†—
  3. Paheli Bhattacharya, Shounak Paul, Kripabandhu Ghosh, Saptarshi Ghosh, and Adam Wyner. 2019. Identification of rhetorical roles of sentences in indian legal judgments. In Legal Knowledge and Information Systems, pages 3–12. IOS Press. β†—
  4. Ilias Chalkidis, Manos Fergadiotis, Prodromos Malakasiotis, Nikolaos Aletras, and Ion Androutsopoulos. 2020. Legal-bert: The muppets straight out of law school. In Findings of the Association for Computational Linguistics: EMNLP 2020, pages 2898–2904. find our card
  5. Ashish Chouhan and Michael Gertz. 2024. Lex-Drafter: Terminology drafting for legislative documents using retrieval augmented generation. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), pages 10448–10458, Torino, Italia. ELRA and ICCL. find
  6. Jacob Cohen. 1960. A coefficient of agreement for nominal scales. Educational and psychological measurement, 20(1):37–46. β†—
  7. 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. β†—
  8. Joseph L Fleiss. 1971. Measuring nominal scale agreement among many raters. Psychological bulletin, 76(5):378. β†—
  9. Abe Bohan Hou, Orion Weller, Guanghui Qin, Eugene Yang, Dawn Lawrie, Nils Holzenberger, Andrew Blair-Stanek, and Benjamin Van Durme. 2025. CLERC: A dataset for U. S. legal case retrieval and retrieval-augmented analysis generation. In Findings of the Association for Computational Linguistics: NAACL 2025, pages 7898– 7913, Albuquerque, New Mexico. Association for Computational Linguistics. find
  10. Klaus Krippendorff. 2018. Content analysis: An introduction to its methodology. Sage publications. β†—
  11. Xiaolin Li, Lei Huang, Yifan Zhou, and Changcheng Shao. 2021. Tst-gan: A legal document generation model based on text style transfer. In 2021 4th International Conference on Robotics, Control and Automation Engineering (RCAE), pages 90–93. IEEE. find
  12. Chin-Yew Lin. 2004. ROUGE: A package for automatic evaluation of summaries. In Text Summarization Branches Out, pages 74–81, Barcelona, Spain. Association for Computational Linguistics. find
  13. Yang Liu, Dan Iter, Yichong Xu, Shuohang Wang, Ruochen Xu, and Chenguang Zhu. 2023. G-eval: NLG evaluation using gpt-4 with better human alignment. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 2511–2522, Singapore. Association for Computational Linguistics. find
  14. Vijit Malik, Rishabh Sanjay, Shouvik Kumar Guha, Angshuman Hazarika, Shubham Kumar Nigam, Arnab Bhattacharya, and Ashutosh Modi. 2022. Semantic segmentation of legal documents via rhetorical roles. In Proceedings of the Natural Legal Language Processing Workshop 2022, pages 153–171, Abu Dhabi, United Arab Emirates (Hybrid). Association for Computational Linguistics. find
  15. Vijit Malik, Rishabh Sanjay, Shubham Kumar Nigam, Kripabandhu Ghosh, Shouvik Kumar Guha, Arnab Bhattacharya, and Ashutosh Modi. 2021. ILDC for CJPE: Indian legal documents corpus for court judgment prediction and explanation. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 4046–4062, Online. Association for Computational Linguistics. find
  16. Masha Medvedeva and Pauline Mcbride. 2023. Legal judgment prediction: If you are going to do it, do it right. In Proceedings of the Natural Legal Language Processing Workshop 2023, pages 73–84. find
  17. Gianluca Moro, Nicola Piscaglia, Luca Ragazzi, and Paolo Italiani. 2024. Multi-language transfer learning for low-resource legal case summarization. Artificial Intelligence and Law, 32(4):1111– 1139. β†—
  18. Gianluca Moro and Luca Ragazzi. 2022. Semantic self-segmentation for abstractive summarization of long documents in low-resource regimes. In Proceedings of the AAAI conference on artificial intelligence, volume 36, pages 11085–11093. find
  19. Shubham Nigam, Anurag Sharma, Danush Khanna, Noel Shallum, Kripabandhu Ghosh, and Arnab Bhattacharya. 2024. Legal judgment reimagined: PredEx and the rise of intelligent AI interpretation in Indian courts. In Findings of the Association for Computational Linguistics ACL 2024, pages 4296–4315, Bangkok, Thailand and virtual meeting. Association for Computational Linguistics. find
  20. Shubham Kumar Nigam, Deepak Patnaik Balaramamahanthi, Shivam Mishra, Noel Shallum, Kripabandhu Ghosh, and Arnab Bhattacharya. 2025. NYAYAANUMANA and INLEGALLLAMA: The largest Indian legal judgment prediction dataset and specialized language model for enhanced decision analysis. In Proceedings of the 31st International Conference on Computational Linguistics, pages 11135–11160, Abu Dhabi, UAE. Association for Computational Linguistics. find
  21. Shubham Kumar Nigam, Navansh Goel, and Arnab Bhattacharya. 2022. nigam@ coliee-22: Legal case retrieval and entailment using cascading of lexical and semantic-based models. In JSAI International Symposium on Artificial Intelligence, pages 96–108. Springer. find
  22. Shubham Kumar Nigam, Shubham Kumar Mishra, Ayush Kumar Mishra, Noel Shallum, and Arnab Bhattacharya. 2023. Legal question-answering in the indian context: Efficacy, challenges, and potential of modern ai models. arXiv preprint arXiv:2309.14735. β†—
  23. Vasile PΔƒis , , Maria Mitrofan, Carol-Luca Gasan, Vlad Coneschi, and Alexandru Ianov. 2021. Named entity recognition in the romanian legal domain. In Proceedings of the natural legal language processing workshop 2021, pages 9–18. find
  24. Rithik Raj Pandey, Sarthak Khandelwal, Satyam Srivastava, Yash Triyar, and Muquitha Almas. 2024. LEGALSEVA - AI-powered legal documentation assistant. International Research Journal of Modernization in Engineering, Technology and Science, 6(3):6418–6423. find
  25. Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. Bleu: a method for automatic evaluation of machine translation. In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, pages 311–318, Philadelphia, Pennsylvania, USA. Association for Computational Linguistics. β†— our card
  26. Atharv Patil, Kartik Bapna, and Ayush Shah. 2024. Legal docgen using ai: Your smart doc generator. International Journal of Novel Research and Development, 9(5):536–543. find
  27. Luca Ragazzi, Gianluca Moro, Stefano Guidi, and Giacomo Frisoni. 2024. Lawsuit: a large expert-written summarization dataset of italian constitutional court verdicts. Artificial Intelligence and Law, pages 1–37. β†—
  28. TYSS Santosh, Apolline Isaia, Shiyu Hong, and Matthias Grabmair. 2024. Hiculr: Hierarchical curriculum learning for rhetorical role labeling of legal documents. arXiv preprint arXiv:2409.18647. β†—
  29. Patrick E Shrout and Joseph L Fleiss. 1979. Intraclass correlations: uses in assessing rater reliability. Psychological bulletin, 86(2):420. β†—
  30. Abhay Shukla, Paheli Bhattacharya, Soham Poddar, Rajdeep Mukherjee, Kripabandhu Ghosh, Pawan Goyal, and Saptarshi Ghosh. 2022. Legal case document summarization: Extractive and abstractive methods and their evaluation. arXiv preprint arXiv:2210.07544. β†—
  31. Takaaki Tateishi, Sachiko Yoshihama, Naoto Sato, and Shin Saito. 2019. Automatic smart contract generation using controlled natural language and template. IBM Journal of Research and Development, 63(2/3):6–1. β†—
  32. Chroma Team. 2023. Chroma: Open-source embedding database. Accessed: July 2025. find
  33. Gemma Team, Aishwarya Kamath, Johan Ferret, Shreya Pathak, Nino Vieillard, Ramona Merhej, Sarah Perrin, Tatiana Matejovicova, Alexandre Ramé, Morgane Rivière, et al. 2025. Gemma 3 technical report. arXiv preprint arXiv:2503.19786. ↗
  34. Yu Tong, Weiming Tan, Jingzhi Guo, Bingqing Shen, Peng Qin, and Shuaihe Zhuo. 2022. Smart contract generation assisted by ai-based word segmentation. Applied Sciences, 12(9):4773. β†—
  35. Santosh T.y.s.s, Mahmoud Aly, Oana Ichim, and Matthias Grabmair. 2025a. LexGenie: Automated generation of structured reports for European court of human rights case law. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track), pages 672–683, Vienna, Austria. Association for Computational Linguistics. find
  36. Santosh T.y.s.s, Youssef Farag, and Matthias Grabmair. 2025b. CoPERLex: Content planning with event-based representations for legal case summarization. In Findings of the Association for Computational Linguistics: NAACL 2025, pages 1016–1032, Albuquerque, New Mexico. Association for Computational Linguistics. β†—
  37. Santosh T.y.s.s and Elvin Quero Hernandez. 2025. LexKeyPlan: Planning with keyphrases and retrieval augmentation for legal text generation: A case study on European court of human rights cases. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 425–436, Vienna, Austria. Association for Computational Linguistics. find
  38. Oleg V. Vasilyev, Vedant Dharnidharka, and John Bohannon. 2020. Fill in the BLANC: human-free quality estimation of document summaries. CoRR, abs/2002.09836. β†—
  39. Shaurya Vats, Atharva Zope, Somsubhra De, Anurag Sharma, Upal Bhattacharya, Shubham Nigam, Shouvik Guha, Koustav Rudra, and Kripabandhu Ghosh. 2023. Llms–the good, the bad or the indispensable?: A use case on legal statute prediction and legal judgment prediction on indian court cases. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 12451–12474. find
  40. Haifeng Wei. 2024. Intelligent legal document generation system and method based on knowledge graph. In Proceedings of the 2024 International Conference on Machine Intelligence and Digital Applications, pages 350–354. find
  41. An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, et al. 2025. Qwen3 technical report. arXiv preprint arXiv:2505.09388. β†— our card
  42. Tianyi Zhang, Varsha Kishore, Felix Wu, Kilian Q. Weinberger, and Yoav Artzi. 2020. Bertscore: Evaluating text generation with BERT. In 8th International Conference on Learning Representations, ICLR 2020, Addis Ababa, Ethiopia, April 26-30, 2020. OpenReview.net. β†—

↑ back to top

This page in Markdown β†’ https://md.research.elamur.ai/free/papers/nigam-2026-structured-legal-document-generation-india/