Title: On large automata processing: towards a high level distributed graph language

Authors: Alpha Mouhamadou Diop; Cheikh Ba

Addresses: LANI, Université Gaston Berger, Saint-Louis, Sénégal ' LANI, Université Gaston Berger, Saint-Louis, Sénégal

Abstract: Large graphs or automata have their data that cannot fit in a single machine, or may take unreasonable time to be processed. We implement with MapReduce and Giraph two algorithms for intersecting and minimising large and distributed automata. We provide some comparative analysis, and the experiment results are depicted in figures. Our work experimentally validates our propositions as long as it shows that our choice, in comparison with MapReduce one, is not only more suitable for graph-oriented algorithms, but also speeds the executions up. This work is one of the first steps of a long-term goal that consists in a high level distributed graph processing language.

Keywords: big data; large graphs and automata; distributed computing; MapReduce; bulk synchronous parallel; BSP.

DOI: 10.1504/IJBDI.2024.138936

International Journal of Big Data Intelligence, 2024 Vol.8 No.2, pp.100 - 109

Received: 21 Mar 2023
Accepted: 20 Aug 2023
Published online: 04 Jun 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article