Gestion efficace de s\'eries temporelles en P2P: Application \`a l'analyse technique et l'\'etude des objets mobiles
Georges Gardarin (PRISM), Benjamin Nguyen (PRISM), Laurent Yeh, (PRISM), Karine Zeitouni (PRISM), Bogdan Butnaru (PRISM), Iulian Sandu-Popa, (PRISM)

TL;DR
This paper introduces a generic model for managing time series data, demonstrating its application in stock investment and ecological transport, and proposes a P2P system for scalable sharing and analysis.
Contribution
It presents a novel generic time series model supporting various operations and a P2P implementation for efficient distributed management and analysis.
Findings
Efficient handling of large time series on distributed systems.
Benchmark results show challenges of classical PCs for large datasets.
P2P approach improves scalability and sharing of time series data.
Abstract
In this paper, we propose a simple generic model to manage time series. A time series is composed of a calendar with a typed value for each calendar entry. Although the model could support any kind of XML typed values, in this paper we focus on real numbers, which are the usual application. We define basic vector space operations (plus, minus, scale), and also relational-like and application oriented operators to manage time series. We show the interest of this generic model on two applications: (i) a stock investment helper; (ii) an ecological transport management system. Stock investment requires window-based operations while trip management requires complex queries. The model has been implemented and tested in PHP, Java, and XQuery. We show benchmark results illustrating that the computing of 5000 series of over 100.000 entries in length - common requirements for both applications -…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsTime Series Analysis and Forecasting · Advanced Database Systems and Queries · Data Management and Algorithms
