Skip to Main content Skip to Navigation

Imitates

Oleksandra Levchenko 1 Djamel-Edine Yagoubi 1 Reza Akbarinia 1 Florent Masseglia 1 Boyan Kolev 1 Dennis Shasha 2 Themis Palpanas 3, 4 Patrick Valduriez 1
1 ZENITH - Scientific Data Management
LIRMM - Laboratoire d'Informatique de Robotique et de Microélectronique de Montpellier, CRISAM - Inria Sophia Antipolis - Méditerranée
2 CIMS - Courant Institute of Mathematical Sciences [New York]
3 UPD5 - Université Paris Descartes - Paris 5
4 LIPADE - EA 2517 - Laboratoire d'Informatique Paris Descartes
Abstract : Imitates is a complete solution for similarity search over very large time series datasets (Terabytes). It implements two efficient approaches for time series indexing and querying: DPiSAX and parSketch. DPiSAX is a parallel solution developed to construct iSAX-based index over billions of time series by making the most of the parallel environment by carefully distributing the work load. The other approach, i.e., parSketch, is based on sketches / random projections to efficiently perform both the parallel indexing of large sets of time series and a similarity search on them. Our solutions take advantage of the computing power of distributed systems by using the Spark parallel framework. The experiments illustrate the high performance of our index construction solution with an indexing time of less than 2 hours for more than 1 billion time series, while the baseline centralized algorithm needs more than 5 days.
Document type :
Software
Complete list of metadatas
https://hal.inria.fr/hal-02095640
Contributor : Reza Akbarinia <>
Submitted on : Wednesday, April 10, 2019 - 3:56:05 PM
Last modification on : Tuesday, December 8, 2020 - 6:20:02 PM

Metadata

Keywords : iSax Sketch Parallel Similarity search Time series

softwareLicence
Apache License 2.0

codeRepository
http://imitates.gforge.inria.fr/

Collections

ZENITH | CNRS | LIRMM | UNIV-MONTPELLIER | LIPADE | UNIV-PARIS5 | MIPS | INRIA | USPC | UNIV-PARIS | UP-SCIENCES

Citation

Oleksandra Levchenko, Djamel-Edine Yagoubi, Reza Akbarinia, Florent Masseglia, Boyan Kolev, et al.. Imitates. 2019. ⟨hal-02095640⟩

Share

Metrics

Record views

209