Elamur Research⌕
☰
home / papers / 2026 / ODUTQA-MDC: A Task for Open-Domain Underspecified Tabular QA with Multi-turn Dialogue-based Clarification

ODUTQA-MDC: A Task for Open-Domain Underspecified Tabular QA with Multi-turn Dialogue-based Clarification

Paper · 2026 · Zhensheng Wang, ZhanTeng Lin, Wenmian Yang et al.

Summary

Tabular QA systems built on LLMs often fail in open-domain settings, where the system must find the right table in a large database and user questions are vague, contain typos, or omit key details. Earlier work mostly stayed in closed-domain setups, only flagged ambiguity, or used fixed scripted dialogues. The authors define ODUTQA-MDC, a task in which a system must detect underspecification and resolve it by asking a simulated user follow-up questions before retrieving tables and generating SQL. They group ambiguity by the SQL clause it affects: scope (FROM), intent (SELECT), condition (WHERE), and mixed.

The benchmark has three parts. ODUTQA is a dataset derived from RETQA. It comes with fine-grained detection labels and a dynamic user-simulator interface. The authors also propose MAIC-TQA, a multi-agent framework that uses SQL execution feedback to drive iterative clarification. Across several LLMs, MAIC-TQA beats the SLUTQA baseline on table retrieval and on SQL execution metrics. The results suggest that interactive clarification is important for realistic tabular QA.

Key points

  • ODUTQA dataset: 209 tables and 25,105 QA pairs from the property, land auction, and real estate finance domains, stored in PostgreSQL.
    • The tables are 102 property tables split by city, 103 land auction tables, and 4 yearly finance tables (2019–2022).
    • The data was built by adapting 76 of RETQA's 90 templates into 222 extended templates.
  • How ambiguity was injected:
    • Templates were restructured to create SELECT ambiguity and missing-scope ambiguity.
    • Values were modified to create typo and mismatch errors in FROM.
    • Proper nouns were replaced with LLM-generated abbreviations to create WHERE ambiguity.
  • MAIC-TQA framework: four agents work together: an SLU module, a Scope Validator, a Table Retrieval agent, and a SQL Generation and Validation agent. The setup forms a "detection–clarification–redetection" loop.
  • Results:
    • On Doubao Seed 1.6, ECR improves over SLUTQA by 12.29% in the fixed scenario and 19.04% in the dynamic scenario.
    • On Qwen3 32B, EA improves by 33.24% (fixed) and 33.69% (dynamic).
    • Detecting mixed underspecification remains the hardest subtask.
  • Limitations:
    • No GROUP BY or HAVING queries.
    • The user simulator is idealized.
    • Only the real estate domain is 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

🎓Beijing Normal UniversityBeijing · China15 works · 3,177 citations🎓Beijing Normal-Hong Kong Baptist UniversityZhuhai · China3 works · 536 citations🎓Beijing Normal University, ZhuhaiZhuhai · China3 works

 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

🧭 Topics 2Research subjects and languages the work is about.Dialogue systems · 624 worksQuestion answering · 441 works

Cite

Cite as Zhensheng Wang, ZhanTeng Lin, Wenmian Yang, Kun Zhou, Yiquan Zhang and Weijia Jia. 2026. ODUTQA-MDC: A Task for Open-Domain Underspecified Tabular QA with Multi-turn Dialogue-based Clarification. 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.499

@inproceedings{wang2026odutqa,
  title = {{ODUTQA-MDC: A Task for Open-Domain Underspecified Tabular QA with Multi-turn Dialogue-based Clarification}},
  author = {Zhensheng Wang, ZhanTeng Lin, Wenmian Yang, Kun Zhou, Yiquan Zhang and Weijia Jia},
  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.499},
  url = {https://doi.org/10.18653/v1/2026.acl-long.499}
}

References

28 references, 8 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Adithya Bhaskar, Tushar Tomar, Ashutosh Sathe, and Sunita Sarawagi. 2023. Benchmarking and improving text-to-sql generation under ambiguity. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, EMNLP 2023, Singapore, December 6-10, 2023, pages 7053–7074. Associa… find
  2. Ziyang Chen, Jinzhi Liao, and Xiang Zhao. 2023. Multi-granularity temporal question answering over knowledge graphs. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), ACL 2023, Toronto, Canada, July 9-14, 2023, pages 11378–11392. Asso… find
  3. An-Chieh Cheng, Hongxu Yin, Yang Fu, Qiushan Guo, Ruihan Yang, Jan Kautz, Xiaolong Wang, and Sifei Liu. 2024. Spatialrgpt: Grounded spatial reasoning in vision-language models. In Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2024, N… find
  4. Lizhi Cheng, Wenmian Yang, and Weijia Jia. 2023. A scope sensitive and result attentive model for multi-intent spoken language understanding. In Proceedings of the AAAI Conference on Artificial Intelligence, pages 12691–12699. AAAI Press. find
  5. 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,…
  6. Mingwen Dong, Nischal Ashok Kumar, Yiqun Hu, Anuj Chauhan, Chung-Wei Hang, Shuaichen Chang, Lin Pan, Wuwei Lan, Henghui Zhu, Jiarong Jiang, Patrick Ng, and Zhiguo Wang. 2025. PRACTIQ: A practical conversational text-to-sql dataset with ambiguous and unanswerable queries. In Proceedings of the 2025 C…
  7. Yifan Duan, Yihong Tang, Kehai Chen, Liqiang Nie, and Min Zhang. 2025. ORPP: Self-optimizing role-playing prompts to enhance language model capabilities. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 28585–28600, Suzhou, China. Association for Compu… find
  8. Xi Fang, Weijie Xu, Fiona Anting Tan, Jiani Zhang, Ziqing Hu, Yanjun Qi, Scott Nickleach, Diego Socol-insky, Srinivasan H. Sengamedu, and Christos Faloutsos. 2024. Large language models(llms) on tabular data: Prediction, generation, and understanding - A survey. CoRR, abs/2402.17944. — [arXiv:2402.17944](https://arxiv.org/abs/2402.17944) find
  9. Team GLM, Aohan Zeng, Bin Xu, and Bowen Wang et al. 2024. Chatglm: A family of large language models from glm-130b to glm-4 all tools. Preprint, arXiv:2406.12793.
  10. Kezhi Kong, Jiani Zhang, Zhengyuan Shen, Balasubramaniam Srinivasan, Chuan Lei, Christos Faloutsos, Huzefa Rangwala, and George Karypis. 2024. Opentab: Advancing large language models as open-domain table reasoners. CoRR, abs/2402.14361. — [arXiv:2402.14361](https://arxiv.org/abs/2402.14361) find
  11. Sunjun Kweon, Yeonsu Kwon, Seonhee Cho, Yohan Jo, and Edward Choi. 2023. Open-wikitable : Dataset for open domain question answering with complex reasoning over table. In Findings of the Association for Computational Linguistics: ACL 2023, Toronto, Canada, July 9-14, 2023, pages 8285–8297. Associati… find
  12. Zeming Liu, Haifeng Wang, Zeyang Lei, Zheng-Yu Niu, Hua Wu, and Wanxiang Che. 2025. Towards few-shot mixed-type dialogue generation. Sci. China Inf. Sci., 68(2). find
  13. Libo Qin, Qiguang Chen, Xiachong Feng, Yang Wu, Yongheng Zhang, Yinghui Li, Min Li, Wanxiang Che, and Philip S. Yu. 2026. Large language models meet NLP: a survey. Frontiers Comput. Sci., 20(11):2011361. find
  14. Libo Qin, Qiguang Chen, Jingxuan Zhou, Jin Wang, Hao Fei, Wanxiang Che, and Min Li. 2025. Divide-solve-combine: An interpretable and accurate prompting framework for zero-shot multi-intent detection. In AAAI-25, Sponsored by the Association for the Advancement of Artificial Intelligence, February 25… find
  15. Libo Qin, Tailu Liu, Wanxiang Che, Bingbing Kang, Sendong Zhao, and Ting Liu. 2021. A co-interactive transformer for joint slot filling and intent detection. In ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 8193–8197. IEEE. find
  16. Daking Rai, Bailin Wang, Yilun Zhou, and Ziyu Yao. 2023. Improving generalization in language model-based text-to-sql semantic parsing: Two simple semantic boundary-based techniques. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers),… find
  17. Tonghui Ren, Yuankai Fan, Zhenying He, Ren Huang, Jiaqi Dai, Can Huang, Yinan Jing, Kai Zhang, Yifan Yang, and X. Sean Wang. 2024. PURPLE: making a large language model a better SQL writer. In 40th IEEE International Conference on Data Engineering, ICDE 2024, Utrecht, The Netherlands, May 13-16, 202… find
  18. Stephen E. Robertson and Hugo Zaragoza. 2009. The probabilistic relevance framework: BM25 and beyond. Found. Trends Inf. Retr., 3(4):333–389. find
  19. Irina Saparina and Mirella Lapata. 2024. AMBROSIA: A benchmark for parsing ambiguous questions into database queries. In Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2024, NeurIPS 2024, Vancouver, BC, Canada, December 10 - 15, 2024. find
  20. Michael Stonebraker, Lawrence A. Rowe, and Michael Hirohama. 2019. The implementation of POST-GRES. In Michael L. Brodie, editor, Making Databases Work: the Pragmatic Wisdom of Michael Stonebraker, volume 22 of ACM Books, pages 519– 559. ACM / Morgan & Claypool. find
  21. Kimi Team, Yifan Bai, Yiping Bao, and et al. 2025. Kimi k2: Open agentic intelligence. Preprint, arXiv:2507.20534.
  22. Zhensheng Wang, Wenmian Yang, Kun Zhou, Yiquan Zhang, and Weijia Jia. 2025. RETQA: A large-scale open-domain tabular question answering dataset for real estate sector. In AAAI-25, Sponsored by the Association for the Advancement of Artificial Intelligence, February 25 - March 4, 2025, Philadelphia,… find
  23. Bowen Xing, Libo Qin, Zhihong Zhu, Zhou Yu, and Ivor W Tsang. 2025. Dxa-net: Dual-task cross-lingual alignment network for zero-shot cross-lingual spoken language understanding. IEEE Transactions on Pattern Analysis and Machine Intelligence. find
  24. An Yang, Anfeng Li, Baosong Yang, and et al. 2025. Qwen3 technical report. CoRR, abs/2505.09388.
  25. Linhao Ye, Lang Yu, Zhikai Lei, Qin Chen, Jie Zhou, and Liang He. 2025. Optimizing question semantic space for dynamic retrieval-augmented multi-hop question answering. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), ACL 2025, Vienn… find
  26. Tao Yu, Rui Zhang, Kai Yang, Michihiro Yasunaga, Dongxu Wang, Zifan Li, James Ma, Irene Li, Qingning Yao, Shanelle Roman, Zilin Zhang, and Dragomir R. Radev. 2018. Spider: A large-scale human-labeled dataset for complex and cross-domain semantic parsing and text-to-sql task. In Proceedings of the 20…
  27. Lianmin Zheng, Liangsheng Yin, Zhiqiang Xie, Chuyue Sun, Jeff Huang, Cody Hao Yu, Shiyi Cao, Christos Kozyrakis, Ion Stoica, Joseph E. Gonzalez, Clark W. Barrett, and Ying Sheng. 2024. Sglang: Efficient execution of structured language model programs. In Advances in Neural Information Processing Sys…
  28. Victor Zhong, Caiming Xiong, and Richard Socher. 2017. Seq2sql: Generating structured queries from natural language using reinforcement learning. CoRR, abs/1709.00103.

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wang-2026-odutqa-mdc-task-open-domain/