Elamur Research⌕
☰
home / papers / 2026 / Hemolix.TabGen: Optimized Table Generation from Documents

Hemolix.TabGen: Optimized Table Generation from Documents

Paper · 2026 · Gyanendra Shrestha, Todor Ivanov, Karthik Vemireddy et al.

Summary

Turning heterogeneous documents in enterprise and medical data lakes into structured tables remains difficult. Classical information extraction methods such as Text-to-Table require domain-specific labeled data and handle only short texts and simple relational outputs. LLM-based systems such as LOTUS, Doctopus and TabAgent work zero-shot, but they cannot produce Bi-dimensional tables with hierarchical metadata, and LLMs in general hallucinate and incur high costs.

Hemolix.TabGen is an unsupervised, domain-agnostic system that reads whole multi-page documents. It outputs relational or Bi-dimensional tables, either with no schema or by following a user-supplied template. Its optimizer builds execution plans from a set of strategies, including Tab2JSON, Evidence Grounding, Multi-step extraction, KG augmentation and LLM-based table validation. A profiler draws on Document Profiles and LLM Profiles to pick the model that best matches user priorities for accuracy, cost or speed.

Across four public multi-domain datasets, the system raised Average Precision by up to 30% over vanilla LLMs. This makes it relevant to practical settings, such as cancer research at Moffitt, where preserving complex table hierarchies is essential.

Key points

  • The evaluation covers four public corpora, including the MCR (1,500 publications) and SCBC (320 publications) oncology collections; in these, about 80% of tables are Bi-dimensional and 20% relational.
  • Tab2JSON raised Metadata Precision from 0% to 100% on hierarchical tables. Evidence Grounding improved AP by up to 21.3%, and KG augmentation by up to 10%.
  • Multi-step extraction lowered hallucination rates from 5.7% to 3.1% for GPT-4 and from 8.2% to 4.8% for Claude. Document-wide comprehension improved BERTScore by up to 8.9%.
  • Automated LLM-based table validation reached 82% average precision, close to the 85% obtained by manual checking.
  • The profiler uses three XGBoost models trained on 1,500 execution records, with F1 of 0.98 for accuracy, 0.91 for cost and 0.90 for latency. Compared with static model selection, it gave 8–12% higher accuracy or 10–15% lower cost.

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

 Fields, topics and concepts

📚 Shelves 5The 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 worksComputer Vision and Pattern Recognition · 2,869 works

Cite

Cite as Gyanendra Shrestha, Todor Ivanov, Karthik Vemireddy, Anna Pyayt and Michael Gubanov. 2026. Hemolix.TabGen: Optimized Table Generation from Documents. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.acl-industry.73

@inproceedings{shrestha2026hemolix,
  title = {{Hemolix.TabGen: Optimized Table Generation from Documents}},
  author = {Gyanendra Shrestha, Todor Ivanov, Karthik Vemireddy, Anna Pyayt and Michael Gubanov},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-industry.73},
  url = {https://doi.org/10.18653/v1/2026.acl-industry.73}
}

References

68 references, 13 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Eric Anderson, Jonathan Fritz, Austin Lee, Bohou Li, Mark Lindblad, Henry Lindeman, Alex Meyer, Parth Parmar, Tanvi Ranade, Mehul A. Shah, Benjamin Sowell, Dan Tecuci, Vinayak Thapliyal, and Matt Welsh. 2024. The design of an llm-powered unstructured analytics system. find
  2. Simran Arora, Avanika Narayan, Mayee F Chen, Laurel J Orr, Neel Guha, Kush Bhatia, Ines Chami, and Christopher Ré. 2023a. Ask me anything: A simple strategy for prompting language models. In ICLR. find
  3. Simran Arora, Brandon Yang, Sabri Eyuboglu, Avanika Narayan, Andrew Hojel, Immanuel Trummer, and Christopher Ré. 2023b. Language models enable simple systems for generating structured views of heterogeneous data lakes. Proc. VLDB Endow.
  4. Evan Bolyen, Jai Ram Rideout, Matthew R Dillon, Nicholas A Bokulich, Christian C Abnet, Gabriel A Al-Ghalith, Harriet Alexander, Eric J Alm, Mani-mozhiyan Arumugam, Francesco Asnicar, and 1 others. 2019. Reproducible, interactive, scalable and extensible microbiome data science using qiime 2. Nature… find
  5. Michael J Cafarella, Christopher Re, Dan Suciu, Oren Etzioni, and Michele Banko. 2007. Structured querying of web text. In 3rd Biennial Conference on Innovative Data Systems Research (CIDR), Asilomar, California, USA. find
  6. Chengliang Chai, Jiajun Li, Yuhao Deng, Yuanhao Zhong, Ye Yuan, Guoren Wang, and Lei Cao. 2025. Doctopus: Budget-aware structural table extraction from unstructured documents. VLDB Endowment. find
  7. Maitry Chauhan, Anna Pyayt, and Michael N. Gubanov. 2023. Learning topical structured interfaces from medical research literature. In ACM Web Conference 2023, WWW. ACM. find
  8. Kaiwen Chen and Nick Koudas. 2024. Unstructured data fusion for schema and data extraction. Proc. ACM Manag. Data, 2(3):181. find
  9. Tianqi Chen and Carlos Guestrin. 2016. Xgboost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 785–794. find
  10. Defne Circi, Ghazal Khalighinejad, Anlan Chen, Bhuwan Dhingra, and L Brinson. 2024. Extracting materials science data from scientific tables. In ACL Workshop Language+ Molecules. find
  11. E. F. Codd. 1983. A relational model of data for large shared data banks. CACM, 26(1):64–69. find
  12. Zheye Deng, Chunkit Chan, Weiqi Wang, Yuxi Sun, Wei Fan, Tianshi Zheng, Yauwai Yim, and Yangqiu Song. 2024. Text-tuple-table: Towards information integration in text-to-table generation via global tuple extraction. In EMNLP. find
  13. 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,…
  14. Haoyu Dong, Yue Hu, Huailiang Peng, and Yanan Cao. 2025. Tablecoder: Table extraction from text via reliable code generation. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track), ACL 2025, Vienna, Austria, July 27 - August 1, 2025, pa… find
  15. M. Gubanov and A. Pyayt. 2013. Readfast: High-relevance search-engine for big text. In ACM CIKM. find
  16. M. Gubanov and A. Pyayt. 2014. Type-aware web search. In EDBT. find
  17. Michael Gubanov. 2017. Polyfuse: A large-scale hybrid data fusion system. In ICDE. find
  18. Michael Gubanov, Manju Priya, and Maksim Podkorytov. 2017. Cognitivedb: An intelligent navigator for large-scale dark structured data. In WWW. find
  19. Michael Gubanov, Anna Pyayt, and Aleksandra Karo-lak. 2024. Cancerkg.org - A web-scale, interactive, verifiable knowledge graph-llm hybrid for assisting with optimal cancer treatment and care. In CIKM. ACM. find
  20. Michael Gubanov and Linda Shapiro. 2012. Using unified famous objects (ufo) to automate alzheimer’s disease diagnostics. In BIBM. find
  21. Michael N Gubanov, Philip A Bernstein, and Alexander Moshchuk. 2008. Model management engine for data integration with reverse-engineering support. In 2008 IEEE 24th International Conference on Data Engineering, pages 1319–1321. IEEE. find
  22. Michael N Gubanov, Lucian Popa, Howard Ho, Hamid Pirahesh, Jeng-Yih Chang, and Shr-Chang Chen. 2009. Ibm ufo repository: Object-oriented data integration. VLDB. find
  23. Michael N. Gubanov, Michael Stonebraker, and Daniel Bruckner. 2014. Text and structured data fusion in data tamer at scale. In ICDE. find
  24. Honghao Gui, Shuofei Qiao, Jintian Zhang, Hongbin Ye, Mengshu Sun, Lei Liang, Jeff Z Pan, Huajun Chen, and Ningyu Zhang. 2024. Instructie: A bilingual instruction-based information extraction dataset. In ISWC. find
  25. Lei Huang, Weijiang Yu, Weitao Ma, Weihong Zhong, Zhangyin Feng, Haotian Wang, Qianglong Chen, Weihua Peng, Xiaocheng Feng, Bing Qin, and 1 others. 2023. A survey on hallucination in large language models: Principles, taxonomy, challenges, and open questions. ACM TOIS.
  26. Todor Ivanov, Gyanendra Shrestha, Karthik Vemireddy, Anna Pyayt, and Michael Gubanov. 2026. Hemolix.extract.v: Llm-based information extraction for documents with ai-based plan selection. In SIGMOD, New York, NY, USA. ACM. find
  27. Peiwen Jiang, Xinbo Lin, Zibo Zhao, Ruhui Ma, Yvonne Jie Chen, and Jinhua Cheng. 2024. Tkgt: Redefinition and a new way of text-to-table tasks based on real world demands and knowledge graphs augmented llms. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, p… find
  28. Bhimesh Kandibedala, Anna Pyayt, Nickolas Piraino, Chris Caballero, and Michael N. Gubanov. 2023. COVIDKG.ORG - a web-scale COVID-19 interactive, trustworthy knowledge graph, constructed and interrogated for bias using deep-learning. In EDBT. OpenProceedings.org. find
  29. Bhimesh Kandibedala, Gyanendra Shrestha, Anna Pyayt, and Michael Gubanov. 2025. Scalable tabular hierarchical metadata classification in heterogeneous structured large-scale datasets using contrastive learning. ICDE. find
  30. Chunwei Liu, Matthew Russo, Michael Cafarella, Lei Cao, Peter Baille Chen, Zui Chen, Michael Franklin, Tim Kraska, Samuel Madden, and Gerardo Vitagliano. 2024. A declarative system for optimizing ai workloads. find
  31. Rituparna Khan and Michael Gubanov. 2018a. Nested dolls: Towards unsupervised clustering of web tables. In IEEE Big Data. find
  32. Fatemeh Nargesian, Erkang Zhu, Renée J Miller, Ken Q Pu, and Patricia C Arocena. 2019. Data lake management: challenges and opportunities. Proceedings of the VLDB Endowment, 12(12):1986–1989. find
  33. Rituparna Khan and Michael Gubanov. 2018b. Towards unsupervised web tables clustering. In IEEE Big-Data. find
  34. Rituparna Khan and Michael Gubanov. 2020a. Towards tabular embeddings, training the relational models. In IEEE Big Data. find
  35. Rituparna Khan and Michael Gubanov. 2020b. We-blens: Towards interactive large-scale structured data profiling. In CIKM. find
  36. Guillaume Lample, Miguel Ballesteros, Sandeep Subramanian, Kazuya Kawakami, and Chris Dyer. 2016. Neural architectures for named entity recognition. In NAACL-HLT.
  37. Jiale Lao, Yibo Wang, Yufei Li, Jianping Wang, Yunjia Zhang, Zhiyuan Cheng, Wanghu Chen, Mingjie Tang, and Jianguo Wang. 2024. Gptuner: A manual-reading database tuning system via gpt-guided bayesian optimization. Proceedings of the VLDB Endowment. find
  38. Jinhyuk Lee, Wonjin Yoon, Sungdong Kim, Donghyeon Kim, Sunkyu Kim, Chan Ho So, and Jaewoo Kang. 2020. Biobert: a pre-trained biomedical language representation model for biomedical text mining. Bioinform.
  39. Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Veselin Stoyanov, and Luke Zettlemoyer. 2020. BART: denoising sequence-to-sequence pre-training for natural language generation, translation, and comprehension. In ACL.
  40. Sha Li, Heng Ji, and Jiawei Han. 2021. Document-level event argument extraction by conditional generation. In NAACL-HLT. ACL.
  41. Tong Li, Zhihao Wang, Liangying Shao, Xuling Zheng, Xiaoli Wang, and Jinsong Su. 2023. A sequence-to-sequence&set model for text-to-table generation. In ACL. find
  42. Yiming Lin, Madelon Hulsebos, Ruiying Ma, Shreya Shankar, Sepanta Zeigham, Aditya G. Parameswaran, and Eugene Wu. 2024. Towards accurate and efficient document analytics with large language models. find
  43. Ying Lin, Heng Ji, Fei Huang, and Lingfei Wu. 2020. A joint neural model for information extraction with global features. In ACL.
  44. Benjamin Newman, Yoonjoo Lee, Aakanksha Naik, Pao Siangliulue, Raymond Fok, Juho Kim, Daniel S Weld, Joseph Chee Chang, and Kyle Lo. 2024. Arxivdigestables: Synthesizing scientific literature into tables using language models. In EMNLP. find
  45. Jekaterina Novikova, Ondrej Dusek, and Verena Rieser. 2017. The E2E dataset: New challenges for end-to-end generation. In SIGDIAL. ACL.
  46. Liana Patel, Siddharth Jha, Parth Asawa, Melissa Pan, Carlos Guestrin, and Matei Zaharia. 2024. Semantic operators: A declarative model for rich, ai-based analytics over text data. find
  47. Sophie Pavia, Rituparna Khan, Anna Pyayt, and Michael Gubanov. 2021. Towards unveiling dark web structured data. In BigData. IEEE. find
  48. Sophie Pavia, Nickolas Piraino, Kazi Islam, Anna Pyayt, and Michael Gubanov. 2022. Hybrid metadata classification in large-scale structured datasets. J. Data Intell., 3(4). find
  49. Michal Pietruszka, Michal Turski, Lukasz Borchmann, Tomasz Dwojak, Gabriela Nowakowska, Karolina Szyndler, Dawid Jurkiewicz, and Lukasz Garncarek. 2024. Stable: Table generation framework for encoder-decoder models. In EACL. find
  50. Maksim Podkorytov, Dylan Soderman, and Michael N. Gubanov. 2017. Hybrid.poly: An interactive large-scale in-memory analytical polystore. In ICDM Workshops, pages 43–50. IEEE Computer Society. find
  51. Cristobal Romero and Sebastian Ventura. 2013. Data mining in education. Wiley Interdisciplinary Reviews: Data mining and knowledge discovery, 3(1):12–27. find
  52. Sunita Sarawagi. 2008. Information extraction. Foundations and Trends® in Databases. find
  53. Shreya Shankar, Tristan Chambers, Tarak Shah, Aditya G. Parameswaran, and Eugene Wu. 2024. Docetl: Agentic query rewriting and evaluation for complex document processing. find
  54. Gyanendra Shrestha, Chutian Jiang, Sai Akula, Vivek Yannam, Anna Pyayt, and Michael Gubanov. 2025. Tabular embeddings for tables with bi-dimensional hierarchical metadata and nesting. In EDBT. find
  55. Mark Simmons, Daniel Armstrong, Dylan Soderman, and Michael Gubanov. 2017. Hybrid.media: High velocity video ingestion in an in-memory scalable analytical polystore. In IEEE Bigdata. find
  56. Ananya Singha, José Cambronero, Sumit Gulwani, Vu Le, and Chris Parnin. 2023. Tabular representation, noisy operators, and impacts on table structure understanding tasks in llms. arXiv preprint arXiv:2310.10358. — [arXiv:2310.10358](https://arxiv.org/abs/2310.10358) find
  57. Brandon Smock, Rohith Pesala, and Robin Abraham. 2022. Pubtables-1m: Towards comprehensive table extraction from unstructured documents. In CVPR. find
  58. Lianmin Zheng, Wei-Lin Chiang, Ying Sheng, Siyuan Zhuang, Zhanghao Wu, Yonghao Zhuang, Zi Lin, Zhuohan Li, Dacheng Li, Eric Xing, Hao Zhang, Joseph E. Gonzalez, and Ion Stoica. 2023. Judging llm-as-a-judge with mt-bench and chatbot arena. NeurIPS.
  59. Zexuan Zhong and Danqi Chen. 2021. A frustratingly easy approach for entity and relation extraction. In NAACL-HLT. ACL.
  60. Santiago Villasenor, Tom Nguyen, Anusha Kola, Sean Soderman, and Michael Gubanov. 2017. Scalable spam classifier for web tables. In IEEE Big Data. find
  61. Lucy Lu Wang, Kyle Lo, Yoganand Chandrasekhar, Russell Reas, Jiangjiang Yang, Doug Burdick, Darrin Eide, Kathryn Funk, Yannis Katsis, Rodney Michael Kinney, and 1 others. 2020. Cord-19: The covid-19 open research dataset. In Proceedings of the 1st Workshop on NLP for COVID-19 at ACL 2020.
  62. Xingbo Wang, Samantha L Huey, Rui Sheng, Saurabh Mehta, and Fei Wang. 2024. Scidasynth: Interactive structured knowledge extraction and synthesis from scientific literature with large language model. arXiv preprint arXiv:2404.13765. — [arXiv:2404.13765](https://arxiv.org/abs/2404.13765) find
  63. Sam Wiseman, Stuart M. Shieber, and Alexander M. Rush. 2017. Challenges in data-to-document generation. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, EMNLP 2017, Copenhagen, Denmark, September 9-11, 2017, pages 2253–2263. Association for Computational Lin…
  64. Jingfei Wu, Chaoyuan Shen, Qiyan Deng, Yuping Wang, Jiajun Li, Yuhao Deng, and Minghe Yu. 2025. Tabagent: A multi-agent table extraction framework for unstructured documents. VLDB Endowment. ISSN. find
  65. Xueqing Wu, Jiacheng Zhang, and Hang Li. 2022. Text-to-table: A new way of information extraction. In ACL. find
  66. Jing Xie, James B Wendt, Yichao Zhou, Seth Ebner, and Sandeep Tata. 2024. Fieldswap: Data augmentation for effective form-like document extraction. In ICDE. find
  67. Wael MS Yafooz, Siti ZZ Abidin, Nasiroh Omar, and Zanariah Idrus. 2013. Managing unstructured data in relational databases. In 2013 IEEE Conference on Systems, Process & Control (ICSPC), pages 198–203. IEEE. find
  68. Yue Zhang, Yafu Li, Leyang Cui, Deng Cai, Lemao Liu, Tingchen Fu, Xinting Huang, Enbo Zhao, Yu Zhang, Yulong Chen, and 1 others. 2025. Siren’s song in the ai ocean: A survey on hallucination in large language models: Siren’s song in the ai ocean: A survey on hallucination in large language models”.… find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/shrestha-2026-hemolix-tabgen-optimized-table-generation/