Calculation of phrase probabilities for Statistical Machine Translation by using belief functions

Abstract : In this paper, we consider a specific part of statistical machine translation: feature estimation for the translation model. The classical way to estimate these features is based on relative frequencies. In this new approach, we propose to use the concept of belief masses to estimate the phrase translation probabilities. The Belief Function theory has proven to be suitable and adapted for dealing with uncertainties in many domains. We have performed a series of experiments to translate from English into French and from Arabic into English showing that our approach performs, at least as well as and at times better than, the classical approach.
Document type :
Conference papers
Complete list of metadatas

Cited literature [23 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01158098
Contributor : Christophe Servan <>
Submitted on : Friday, May 29, 2015 - 3:11:14 PM
Last modification on : Tuesday, December 19, 2017 - 2:42:03 PM
Long-term archiving on : Monday, April 24, 2017 - 9:36:00 PM

File

C12-2107.pdf
Publisher files allowed on an open archive

Identifiers

  • HAL Id : hal-01158098, version 1

Collections

Citation

Christophe Servan, Simon Petitrenaud. Calculation of phrase probabilities for Statistical Machine Translation by using belief functions. The 24th International Conference on Computational Linguistics (COLING 2012), Dec 2012, Mumbai, India. ⟨hal-01158098⟩

Share

Metrics

Record views

162

Files downloads

57