Affiliation:
1. Mohammed VI Polytech. Univ.
2. Université Paris Cité & FORTH
3. Snowflake Inc.
4. Université Paris Cité & IUF
5. IRDA
Abstract
We propose Hercules, a parallel tree-based technique for exact similarity search on massive disk-based data series collections. We present novel index construction and query answering algorithms that leverage different summarization techniques, carefully schedule costly operations, optimize memory and disk accesses, and exploit the multi-threading and SIMD capabilities of modern hardware to perform CPU-intensive calculations. We demonstrate the superiority and robustness of Hercules with an extensive experimental evaluation against state-of-the-art techniques, using many synthetic and real datasets, and query workloads of varying difficulty. The results show that Hercules performs up to one order of magnitude faster than the best competitor (which is not always the same). Moreover, Hercules is the only index that outperforms the optimized scan on all scenarios, including the hard query workloads on disk-based datasets.
Publisher
Association for Computing Machinery (ACM)
Subject
General Earth and Planetary Sciences,Water Science and Technology,Geography, Planning and Development
Reference71 articles.
1. 2018. Lernaean Hydra Archive. http://www.mi.parisdescartes.fr/~themisp/dsseval/. 2018. Lernaean Hydra Archive. http://www.mi.parisdescartes.fr/~themisp/dsseval/.
2. 2019. Faiss. https://github.com/facebookresearch/faiss/. 2019. Faiss. https://github.com/facebookresearch/faiss/.
3. 2019. Return of Lernaean Hydra Archive. http://www.mi.parisdescartes.fr/~themisp/dsseval2/. 2019. Return of Lernaean Hydra Archive. http://www.mi.parisdescartes.fr/~themisp/dsseval2/.
4. 2022. Hercules Archive. http://www.mi.parisdescartes.fr/~themisp/hercules/. 2022. Hercules Archive. http://www.mi.parisdescartes.fr/~themisp/hercules/.
5. Flexible non-linear predictive models for large-scale wind turbine diagnostics
Cited by
6 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献