FaDE: More Than a Million What-Ifs Per Second
Proceedings of the VLDB Endowment · 2024-12 · 18 (4) · 943-955
DOI: 10.14778/3717755.3717757 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the VLDB Endowment, online erschienen am 20.05.2025. Verfasst von Haneen Mohammed und 7 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 37 Literaturangaben.
Abstract (Originalsprache)
What-if queries are the building blocks for many explanation and analytics applications—sensitivity analysis, hypothetical reasoning, data cleaning, probabilistic databases—that explore how a query's output changes due to input data changes. Their response time is bounded by intervention evaluation latency, which can be in the minute or hours for complex queries and large datasets. FaDE is a compilation engine that uses provenance to evaluate hypothetical deletion and scaling interventions at low latency and high throughput. FaDE forgoes conventional provenance representations as symbolic expressions and leverages their underlying relational structure. This accelerates intervention evaluation on average by 1000× against IVM and 10,000× against prior provenance-based approaches. In addition, FaDE develops a suite of optimizations (e.g., compilation, parallelization, incremental evaluation, sparse representations) that collectively raise evaluation throughput to >1 million interventions per sec—a rate that can brute-force existing applications within 1 s.
Zitieren
APA
Mohammed, H., Yao, A., Summers, C., Zhong, H., Chan, G. Y., Mitra, S., Flokas, L., & Wu, E. (2024). FaDE: More Than a Million What-Ifs Per Second. Proceedings of the VLDB Endowment, 18(4), 943-955. https://doi.org/10.14778/3717755.3717757
Vancouver
Mohammed H, Yao A, Summers C, Zhong H, Chan GY, Mitra S, et al. FaDE: More Than a Million What-Ifs Per Second. Proceedings of the VLDB Endowment. 2024;18(4):943-955. doi:10.14778/3717755.3717757
BibTeX
@article{Mohammed2024,
author = {Mohammed, Haneen and Yao, Alexander and Summers, Charlie and Zhong, Hongbin and Chan, Gromit Yeuk-Yin and Mitra, Subrata and Flokas, Lampros and Wu, Eugene},
title = {FaDE: More Than a Million What-Ifs Per Second},
journal = {Proceedings of the VLDB Endowment},
year = {2024},
volume = {18},
number = {4},
pages = {943--955},
doi = {10.14778/3717755.3717757}
}
EUVIMED