ShadowAQP: Efficient Approximate Group-by and Join Query via Attribute-oriented Sample Size Allocation and Data Generation
Summary: ShadowAQP allocates sample sizes per group-by/join attribute-value combinations and synthesizes table rows via a conditional VAE (with automatic encoding and model updates) to avoid costly raw sampling. With parallel multi-round aggregation, outlier-aware sampling and dimensionality reduction it yields up to 12.8x speedups and ~74% average error reduction versus SOTA. (summarized by gpt-5-mini on Feb 09 2026)
Incoming Non-self Citations Over Time
Authors
- 1. Rong Gu (Nanjing University)
- 2. Han Li (Nanjing University)
- 3. Haipeng Dai (Nanjing University)
- 4. Wenjie Huang (Nanjing University)
- 5. Jie Xue (New York University)
- 6. Meng Li (Nanjing University)
- 7. Jiaqi Zheng (Nanjing University)
- 8. Haoran Cai
- 9. Yihua Huang (Nanjing University)
- 10. Guihai Chen (Nanjing University)
BibTeX Citation
@article{gu_vldb23,
title = {{ShadowAQP: Efficient Approximate Group-by and Join Query via Attribute-oriented Sample Size Allocation and Data Generation}},
author = {Gu, Rong and Li, Han and Dai, Haipeng and Huang, Wenjie and Xue, Jie and Li, Meng and Zheng, Jiaqi and Cai, Haoran and Huang, Yihua and Chen, Guihai},
journal = {PVLDB},
series = {{VLDB} '23},
volume = {16},
number = {13},
pages = {4216--4229},
doi = {10.14778/3625054.3625059},
url = {https://doi.org/10.14778/3625054.3625059},
year = {2023}
}
Incoming Citations (Sorted by Pagerank)
Showing 3 of 3 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 10,185 | A Backend-Agnostic Compiler for Approximate Query Processing with Probabilistic Tensor Algebra | 2026 | SIGMOD | 5.093636e-05 |
| 10,760 | FAAQP: Fast and Accurate Approximate Query Processing based on Bitmap-augmented Sum-Product Network | 2025 | SIGMOD | 5.093636e-05 |
| 11,194 | Enabling Adaptive Sampling for Intra-Window Join: Simultaneously Optimizing Quantity and Quality | 2024 | SIGMOD | 5.093636e-05 |
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 36 of 36 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
Previous
Page 1 / 1
Next