Robustness of Updatable Learning-based Index Advisors against Poisoning Attack
Summary: Introduces PIPA, an opaque-box stress-test framework to evaluate the robustness of updatable learning-based Index Advisors against poisoning attacks without using private data. Probing, injecting, and IABART-based query generation reveal systemic non-robustness: subtle extraneous workloads can demote top indexes and trap IAs in local optima even after fine-tuning. (summarized by gpt-5-nano on Feb 09 2026)
Incoming Non-self Citations Over Time
Authors
- 1. Yihang Zheng (Xiamen University)
- 2. Chen Lin (Shanghai AI Laboratory; Xiamen University)
- 3. Xian Lyu (Xiamen University)
- 4. Xuanhe Zhou (Tsinghua University)
- 5. Guoliang Li (Tsinghua University)
- 6. Tianqing Wang (Huawei)
BibTeX Citation
@inproceedings{zheng_sigmod24,
title = {{Robustness of Updatable Learning-based Index Advisors against Poisoning Attack}},
author = {Zheng, Yihang and Lin, Chen and Lyu, Xian and Zhou, Xuanhe and Li, Guoliang and Wang, Tianqing},
series = {{SIGMOD} '24},
booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
publisher = {Association for Computing Machinery},
doi = {10.1145/3639265},
url = {https://dl.acm.org/doi/10.1145/3639265},
year = {2024}
}
Incoming Citations (Sorted by Pagerank)
Showing 2 of 2 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 9,599 | Understanding Robustness Issues of Updatable Learned Indexes: [Experiments & Analysis] | 2026 | SIGMOD | 5.2492748e-05 |
| 10,267 | Mathematical Foundations of Poisoning Attacks on Linear Regression over Cumulative Distribution Functions | 2026 | SIGMOD | 5.093636e-05 |
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 10 of 10 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
Previous
Page 1 / 1
Next
Semantically Similar Papers
| # | Overall Rank | Paper | Year | Venue |
|---|---|---|---|---|
| 1 | 847 | Benchmarking Learned Indexes | 2021 | VLDB |
| 2 | 5,091 | Budget-aware Index Tuning with Reinforcement Learning | 2022 | SIGMOD |
| 3 | 3,792 | Learned Index: A Comprehensive Experimental Evaluation | 2023 | VLDB |
| 4 | 10,267 | Mathematical Foundations of Poisoning Attacks on Linear Regression over Cumulative Distribution Functions | 2026 | SIGMOD |
| 5 | 8,039 | Algorithmic Complexity Attacks on Dynamic Learned Indexes | 2024 | VLDB |
| 6 | 2,806 | Are Updatable Learned Indexes Ready? | 2022 | VLDB |
| 7 | 6,327 | Breaking It Down: An In-depth Study of Index Advisors | 2024 | VLDB |
| 8 | 4,516 | The Price of Tailoring the Index to Your Data: Poisoning Attacks on Learned Index Structures | 2022 | SIGMOD |
| 9 | 10,378 | High Performance or Low Memory? An Updatable Learned Index Framework for Time-Space Tradeoff | 2026 | SIGMOD |
| 10 | 9,599 | Understanding Robustness Issues of Updatable Learned Indexes: [Experiments & Analysis] | 2026 | SIGMOD |