[ Skip to the content ]

Institute of Formal and Applied Linguistics

at Faculty of Mathematics and Physics, Charles University, Prague, Czech Republic

[ Back to the navigation ]


Year 2016
Type in proceedings
Status published
Language English
Author(s) Kocúr, Viktor Bojar, Ondřej
Title Particle Swarm Optimization Submission for WMT16 Tuning Task
Czech title Příspěvek do optimalizační soutěže WMT16 Tuning Task pomocí metody reje částic
Proceedings 2016: Stroudsburg, PA, USA: WMT 2016 (ACL): Proceedings of the First Conference on Machine Translation (WMT). Volume 2: Shared Task Papers
Pages range 518-524
How published online
URL http://www.statmt.org/wmt16/pdf/W16-2344.pdf
Supported by 2015-2018 H2020-ICT-2014-1-645452 (QT21: Quality Translation 21) 2016-2019 LM2015071 (LINDAT-CLARIN: Institut pro analýzu, zpracování a distribuci lingvistických dat) 2012-2016 PRVOUK P46 (Informatika)
Czech abstract Článek popisuje náš příspěvek do optimalizační soutěže WMT16 Tuning Task. Grid search ve standardní implementaci MERT nahrazujeme optimalizací pomocí metody reje částic.
English abstract This paper describes our submission to the Tuning Task of WMT16. We replace the grid search implemented as part of standard minimum-error rate training (MERT) in the Moses toolkit with a search based on particle swarm optimization (PSO). An older variant of PSO has been previously successfully applied and we now test it in optimizing the Tuning Task model for English-to-Czech translation. We also adapt the method in some aspects to allow for even easier parallelization of the search.
Specialization linguistics ("jazykověda")
Confidentiality default – not confidential
Open access no
DOI http://dx.doi.org/10.18653/v1/w16-2344
Editor(s)* Ondřej Bojar; et al .
ISBN* 978-1-945626-10-4
Address* Stroudsburg, PA, USA
Month* August
Venue* Humboldt University
Publisher* Association for Computational Linguistics
Institution* Association for Computational Linguistics
Creator: Common Account
Created: 9/7/16 8:41 AM
Modifier: Almighty Admin
Modified: 2/25/17 10:07 PM

Content, Design & Functionality: ÚFAL, 2006–2016. Page generated: Fri Sep 22 01:12:14 CEST 2017

[ Back to the navigation ] [ Back to the content ]

100% OpenAIRE compliant