Heterogeneous tabular data query and analysis system based on large language models

Jiajun LI , Xiaojun SHI , Angeline Mary MARCHELLA , Tianze HU , Lianpeng QIAO , Kaiyu LI , Guoren WANG

Front. Comput. Sci. ›› 2027, Vol. 21 ›› Issue (9) : 2109354

PDF (7285KB)
Front. Comput. Sci. ›› 2027, Vol. 21 ›› Issue (9) :2109354 DOI: 10.1007/s11704-026-60020-2
Artificial Intelligence
RESEARCH ARTICLE
Heterogeneous tabular data query and analysis system based on large language models
Author information +
History +
PDF (7285KB)

Abstract

With the increase of digital information, there is a growing demand for efficiently extracting insights and answering queries from tables. However, real-world tables often feature complex structures like merged cells, multi-level headers, and block layouts. These heterogeneous tables pose significant challenges for question-answering systems. Existing research focuses primarily on normalized tables. Even state-of-the-art large language models for tabular data achieve 31%–42% lower accuracy on heterogeneous tables. Current methods either linearize tables into text and lose structural information or adopt graph-based representations that are computationally inefficient and yield suboptimal accuracy. Moreover, these methods typically require extensive manual processing to convert complex tables into specific formats, incurring additional annotation costs. To address these limitations, this paper presents DTI-HTQA (Dual-Tree-Indexing Heterogeneous Table Question Answering), an end-to-end framework for heterogeneous table question answering. It uses a graph-based representation to preserve table semantics, a LLM-driven header recognition algorithm to eliminate manual annotations, and dual-tree indices for intelligent retrieval. Four core operators guide precise cell retrieval, while a chain-of-thought reasoning strategy enhances transparency and accuracy for complex queries. Experiments on HiTab and AIT-QA show that DTI-HTQA achieves top performance in exact match and LLM-evaluated accuracy, approaching or surpassing state-of-the-art methods. The header recognition algorithm achieves an accuracy of 88%, and an upper-bound analysis confirms that this causes only a 2%–4% accuracy gap compared to gold-standard annotations, validating the feasibility of the LLM-based approach. Comprehensive ablation studies demonstrate that each component contributes meaningfully to the overall performance, with the dual-tree index providing the most substantial gains. Cost-efficiency analysis shows that DTI-HTQA achieves 52%–71% lower token consumption than comparable multi-step frameworks while delivering higher accuracy. This study tackles key challenges in heterogeneous table question answering and offers a practical solution by reducing reliance on costly manual annotations, enabling more efficient use of tabular data.

Graphical abstract

Keywords

heterogeneous tables / large language models / tree indexing / header recognition / tabular data question answering

Cite this article

Download citation ▾
Jiajun LI, Xiaojun SHI, Angeline Mary MARCHELLA, Tianze HU, Lianpeng QIAO, Kaiyu LI, Guoren WANG. Heterogeneous tabular data query and analysis system based on large language models. Front. Comput. Sci., 2027, 21 (9) : 2109354 DOI:10.1007/s11704-026-60020-2

登录浏览全文

4963

注册一个新账户 忘记密码

References

[1]

OpenAI. GPT-4 technical report. 2024, arXiv preprint arXiv: 2303.08774

[2]

Touvron H, Martin L, Stone K, Albert P, Almahairi A, et al. Llama 2: open foundation and fine-tuned chat models. 2023, arXiv preprint arXiv: 2307.09288

[3]

DeepSeek-AI. DeepSeek-V3 technical report. 2025, arXiv preprint arXiv: 2412.19437

[4]

Su A, Wang A, Ye C, Zhou C, Zhang G, et al. TableGPT2: a large multimodal model with tabular data integration. 2024, arXiv preprint arXiv: 2411.02059

[5]

Cheng Z, Dong H, Wang Z, Jia R, Guo J, Gao Y, Han S, Lou J G, Zhang D . HiTab: a hierarchical table dataset for question answering and natural language generation. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2022, 1094–1110

[6]

Zha L, Zhou J, Li L, Wang R, Huang Q, et al. TableGPT: towards unifying tables, nature language and commands into one GPT. 2023, arXiv preprint arXiv: 2307.08674

[7]

Li J, Hui B, Cheng R, Qin B, Ma C, Huo N, Huang F, Du W, Si L, Li Y . Graphix-T5: mixing pre-trained transformers with graph-aware layers for text-to-SQL parsing. Proceedings of the 37th AAAI Conference on Artificial Intelligence, 2023, 13076–13084

[8]

Zhang T, Yue X, Li Y, Sun H. . TableLlama: towards open large generalist models for tables.. Proceedings of 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), 2024, 6024–6044

[9]

Ye Y, Hui B, Yang M, Li B, Huang F, Li Y. Large language models are versatile decomposers: decompose evidence and questions for table-based reasoning. 2023, arXiv preprint arXiv: 2301.13808

[10]

Zhao Y, Li Y, Li C, Zhang R. . MultiHiertt: numerical reasoning over multi hierarchical tabular and textual data.. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2022, 6588–6600

[11]

Fang J, Mitra P, Tang Z, Giles C L. . Table header detection and classification.. Proceedings of the 26th AAAI Conference on Artificial Intelligence., 2012, 599–605

[12]

Zhao W X, Zhou K, Li J, Tang T, Dong Z, . et al. A survey of large language models. Frontiers of Computer Science, 2026, 20( 12): 2012627

[13]

Saul L, Pereira F . Aggregate and mixed-order Markov models for statistical language processing. Proceedings of the 2nd Conference on Empirical Methods in Natural Language Processing, 1997,

[14]

Bengio Y, Ducharme R, Vincent P, Janvin C . A neural probabilistic language model. The Journal of Machine Learning Research, 2003, 3: 1137–1155

[15]

Kim Y, Jernite Y, Sontag D, Rush A M . Character-aware neural language models. Proceedings of the 30th AAAI Conference on Artificial Intelligence, 2016, 2741–2749

[16]

Peters M E, Neumann M, Iyyer M, Gardner M, Clark C, Lee K, Zettlemoyer L. . Deep contextualized word representations.. Proceedings of 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers), 2018, 2227–2237

[17]

Wang H, Li J, Wu H, Hovy E, Sun Y . Pre-trained language models and their applications. Engineering, 2023, 25: 51–65

[18]

Xu L, Xie H, Qin S Z J, Tao X, Wang F L. Parameter-efficient fine-tuning methods for pretrained language models: a critical review and assessment. 2023, arXiv preprint arXiv: 2312.12148

[19]

Raffel C, Shazeer N, Roberts A, Lee K, Narang S, Matena M, Zhou Y, Li W, Liu P J . Exploring the limits of transfer learning with a unified text-to-text transformer. Journal of Machine Learning Research, 2020, 21( 140): 1–67

[20]

Brown T B, Mann B, Ryder N, Subbiah M, Kaplan J, . et al. Language models are few-shot learners.. Proceedings of the 34th International Conference on Neural Information Processing Systems, 2020, 1877-1901

[21]

Wei J, Bosma M, Zhao V Y, Guu K, Yu A W, Lester B, Du N, Dai A M, Le Q V. . Finetuned language models are zero-shot learners.. Proceedings of the 10th International Conference on Learning Representations., 2022,

[22]

Chung H W, Hou L, Longpre S, Zoph B, Tay Y, . et al. Scaling instruction-finetuned language models. Journal of Machine Learning Research, 2024, 25( 70): 1–53

[23]

Wei J, Wang X, Schuurmans D, Bosma M, Ichter B, Xia F, Chi E H, Le Q V, Zhou D. . Chain-of-thought prompting elicits reasoning in large language models.. Proceedings of the 36th International Conference on Neural Information Processing Systems., 2022, 1800

[24]

Liu G, Yang J, Wu L. PTab: using the pre-trained language model for modeling tabular data. 2022, arXiv preprint arXiv: 2209.08060

[25]

Yin P, Neubig G, Yih W T, Riedel S. . TaBERT: pretraining for joint understanding of textual and tabular data.. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics., 2020, 8413–8426

[26]

Chen Z, Trabelsi M, Heflin J, Xu Y, Davison B D. . Table search using a deep contextualized language model.. Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval., 2020, 589–598

[27]

Singha A, Cambronero J, Gulwani S, Le V, Parnin C. Tabular representation, noisy operators, and impacts on table structure understanding tasks in LLMs. 2023, arXiv preprint arXiv: 2310.10358

[28]

Sui Y, Zhou M, Zhou M, Han S, Zhang D. . Table meets LLM: can large language models understand structured table data? A benchmark and empirical study.. Proceedings of the 17th ACM International Conference on Web Search and Data Mining., 2024, 645–654

[29]

Herzig J, Nowak P K, Müller T, Piccinno F, Eisenschlos J . TaPas: weakly supervised table parsing via pre-training. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, 2020, 4320–4333

[30]

Jiang Z, Mao Y, He P, Neubig G, Chen W. . OmniTab: pretraining with natural and synthetic data for few-shot table-based question answering.. Proceedings of 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies., 2022, 932–942

[31]

Katsis Y, Chemmengath S, Kumar V, Bharadwaj S, Canim M, Glass M, Gliozzo A, Pan F, Sen J, Sankaranarayanan K, Chakrabarti S. . AIT-QA: question answering dataset over complex tables in the airline industry.. Proceedings of 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: Industry Track., 2022, 305–314

[32]

Chen Y, Jin D, Huang C, Liu J, Lei W. . TRAVEL: tag-aware conversational FAQ retrieval via reinforcement learning.. Proceedings of 2023 Conference on Empirical Methods in Natural Language Processing., 2023, 3861–3872

[33]

Liu T, Wang F, Chen M. . Rethinking tabular data understanding with large language models.. Proceedings of 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), 2024, 450–482

[34]

Zhang Z, Gao Y, Lou J G. . E5: zero-shot hierarchical table analysis using augmented LLMs via explain, extract, execute, exhibit and extrapolate.. Proceedings of 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), 2024, 1244–1258

[35]

Yao S, Zhao J, Yu D, Du N, Shafran I, Narasimhan K R, Cao Y . ReAct: synergizing reasoning and acting in language models. Proceedings of the 11th International Conference on Learning Representations, 2023,

Rights & permissions

Higher Education Press

PDF (7285KB)

Supplementary files

Highlights

588

Accesses

0

Citation

Detail

Sections
Recommended

/

〈 〉