A new method for learning Phrase Based Machine Translation with Multivariate Mutual Information

Cyrine Nasri 1 Kamel Smaïli 2, 1 Chiraz Latiri 3 Yahya Slimani 3
1 PAROLE - Analysis, perception and recognition of speech
Inria Nancy - Grand Est, LORIA - NLPKD - Department of Natural Language Processing & Knowledge Discovery
2 SMarT - Statistical Machine Translation and Speech Modelization and Text
LORIA - NLPKD - Department of Natural Language Processing & Knowledge Discovery
3 URPAH Tunis
URPAH - Unité de Recherche en Programmation Algorithmique et Heuristique
Abstract : Current statistical machine translation systems usually build an initial word-to-word alignments before learning phrase translation pairs. This operation needs so many matching between di erent single words of both considered languages. We propose a new approach for phrase-based machine translation which does not need any word alignments, it is based on inter-lingual triggers determined by Multivariate Mutual Information. This algorithm segments sentences into phrases and nds their alignments simultaneously. The main objective is to build directly valid alignments between source and target phrases. Inspite of the youth of this method, experiments showed that the results are competitive but needs some more e orts in order to overcome the one of state-of-the-art methods.
Document type :
Conference papers
Liste complète des métadonnées

Cited literature [24 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-00727044
Contributor : Kamel Smaïli <>
Submitted on : Saturday, September 1, 2012 - 1:19:09 AM
Last modification on : Tuesday, December 18, 2018 - 4:38:02 PM

File

102.pdf
Files produced by the author(s)

Licence


Distributed under a Creative Commons Attribution - NonCommercial - NoDerivatives 4.0 International License

Identifiers

  • HAL Id : hal-00727044, version 1

Collections

Citation

Cyrine Nasri, Kamel Smaïli, Chiraz Latiri, Yahya Slimani. A new method for learning Phrase Based Machine Translation with Multivariate Mutual Information. The 8th International Conference on Natural Language Processing and Knowledge Engineering - NLP-KE'12, Sep 2012, HuangShan, China. ⟨hal-00727044⟩

Share

Metrics

Record views

630

Files downloads

175