Replacing Multi-Step Assembly of Data Preparation Pipelines with One-Step LLM Pipeline Generation for Table QA
Summary: Operation-R1 trains lightweight LLMs with verifiable rewards to generate TQA data-preparation pipelines in one inference step, replacing costly multi-call assembly. Operation merging and adaptive rollback improve robustness while delivering higher accuracy, 79% compression, and 2.2× lower cost. (summarized by gpt-5.6-luna on Aug 17 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Fengyu Li (Zhejiang University)
- 2. Junhao Zhu (Zhejiang University)
- 3. Kaishi Song (Zhejiang University)
- 4. Zhongming Yao (Northeastern University)
- 5. Tianyi Li (Aalborg University)
- 6. Lu Chen (Zhejiang University)
- 7. Christian S. Jensen (Aalborg University)
BibTeX Citation
@article{li_vldb26,
title = {{Replacing Multi-Step Assembly of Data Preparation Pipelines with One-Step LLM Pipeline Generation for Table QA}},
author = {Li, Fengyu and Zhu, Junhao and Song, Kaishi and Yao, Zhongming and Li, Tianyi and Chen, Lu and Jensen, Christian S.},
journal = {PVLDB},
series = {{VLDB} '26},
volume = {19},
number = {9},
pages = {2372--2384},
doi = {10.14778/3819518.3819557},
url = {https://doi.org/10.14778/3819518.3819557},
year = {2026}
}Incoming Citations (Sorted by Pagerank)
Showing 0 of 0 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 3 of 3 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 1,683 | ReAcTable: Enhancing ReAct for Table Question Answering | 2024 | VLDB | 9.8822754e-05 |
| 3,423 | AutoTQA: Towards Autonomous Tabular Question Answering through Multi-Agent Large Language Models | 2024 | VLDB | 7.313499e-05 |
| 7,270 | AutoPrep: Natural Language Question-Aware Data Preparation with a Multi-Agent Framework | 2025 | VLDB | 5.5694935e-05 |
Previous
Page 1 / 1
Next