Are Large Language Models Ready for Multi-Turn Tabular Data Analysis?

Jinyang Li, Nan Huo, Yan Gao, Jiayi Shi, Yingxiu Zhao, Ge Qu, Bowen Qin, Yurong Wu, Xiaodong Li, Chenhao Ma, Jian-Guang Lou, Reynold Cheng
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:34795-34835, 2025.

Abstract

Conversational Tabular Data Analysis, a collaboration between humans and machines, enables real-time data exploration for informed decision-making. The challenges and costs of collecting realistic conversational logs for tabular data analysis hinder comprehensive quantitative evaluation of Large Language Models (LLMs) in this task. To mitigate this issue, we introduce CoTA, a new benchmark to evaluate LLMs on conversational tabular data analysis. CoTA contains 1013 conversations, covering 4 practical scenarios: Normal, Action, Private, and Private Action. Notably, CoTA is constructed by an economical multi-agent environment, Decision Company, with few human efforts. This environment ensures efficiency and scalability of generating new conversational data. Our comprehensive study, conducted by data analysis experts, demonstrates that Decision Company is capable of producing diverse and high-quality data, laying the groundwork for efficient data annotation. We evaluate popular and advanced LLMs in CoTA, which highlights the challenges of conversational tabular data analysis. Furthermore, we propose Adaptive Conversation Reflection (ACR), a self-generated reflection strategy that guides LLMs to learn from successful histories. Experiments demonstrate that ACR can evolve LLMs into effective conversational data analysis agents, achieving a relative performance improvement of up to 35.14%.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-li25aj, title = {Are Large Language Models Ready for Multi-Turn Tabular Data Analysis?}, author = {Li, Jinyang and Huo, Nan and Gao, Yan and Shi, Jiayi and Zhao, Yingxiu and Qu, Ge and Qin, Bowen and Wu, Yurong and Li, Xiaodong and Ma, Chenhao and Lou, Jian-Guang and Cheng, Reynold}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {34795--34835}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/li25aj/li25aj.pdf}, url = {https://proceedings.mlr.press/v267/li25aj.html}, abstract = {Conversational Tabular Data Analysis, a collaboration between humans and machines, enables real-time data exploration for informed decision-making. The challenges and costs of collecting realistic conversational logs for tabular data analysis hinder comprehensive quantitative evaluation of Large Language Models (LLMs) in this task. To mitigate this issue, we introduce CoTA, a new benchmark to evaluate LLMs on conversational tabular data analysis. CoTA contains 1013 conversations, covering 4 practical scenarios: Normal, Action, Private, and Private Action. Notably, CoTA is constructed by an economical multi-agent environment, Decision Company, with few human efforts. This environment ensures efficiency and scalability of generating new conversational data. Our comprehensive study, conducted by data analysis experts, demonstrates that Decision Company is capable of producing diverse and high-quality data, laying the groundwork for efficient data annotation. We evaluate popular and advanced LLMs in CoTA, which highlights the challenges of conversational tabular data analysis. Furthermore, we propose Adaptive Conversation Reflection (ACR), a self-generated reflection strategy that guides LLMs to learn from successful histories. Experiments demonstrate that ACR can evolve LLMs into effective conversational data analysis agents, achieving a relative performance improvement of up to 35.14%.} }
Endnote
%0 Conference Paper %T Are Large Language Models Ready for Multi-Turn Tabular Data Analysis? %A Jinyang Li %A Nan Huo %A Yan Gao %A Jiayi Shi %A Yingxiu Zhao %A Ge Qu %A Bowen Qin %A Yurong Wu %A Xiaodong Li %A Chenhao Ma %A Jian-Guang Lou %A Reynold Cheng %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-li25aj %I PMLR %P 34795--34835 %U https://proceedings.mlr.press/v267/li25aj.html %V 267 %X Conversational Tabular Data Analysis, a collaboration between humans and machines, enables real-time data exploration for informed decision-making. The challenges and costs of collecting realistic conversational logs for tabular data analysis hinder comprehensive quantitative evaluation of Large Language Models (LLMs) in this task. To mitigate this issue, we introduce CoTA, a new benchmark to evaluate LLMs on conversational tabular data analysis. CoTA contains 1013 conversations, covering 4 practical scenarios: Normal, Action, Private, and Private Action. Notably, CoTA is constructed by an economical multi-agent environment, Decision Company, with few human efforts. This environment ensures efficiency and scalability of generating new conversational data. Our comprehensive study, conducted by data analysis experts, demonstrates that Decision Company is capable of producing diverse and high-quality data, laying the groundwork for efficient data annotation. We evaluate popular and advanced LLMs in CoTA, which highlights the challenges of conversational tabular data analysis. Furthermore, we propose Adaptive Conversation Reflection (ACR), a self-generated reflection strategy that guides LLMs to learn from successful histories. Experiments demonstrate that ACR can evolve LLMs into effective conversational data analysis agents, achieving a relative performance improvement of up to 35.14%.
APA
Li, J., Huo, N., Gao, Y., Shi, J., Zhao, Y., Qu, G., Qin, B., Wu, Y., Li, X., Ma, C., Lou, J. & Cheng, R.. (2025). Are Large Language Models Ready for Multi-Turn Tabular Data Analysis?. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:34795-34835 Available from https://proceedings.mlr.press/v267/li25aj.html.

Related Material