New Confidence Measures for Statistical Machine Translation - Archive ouverte HAL Access content directly
Conference Papers Year : 2009

New Confidence Measures for Statistical Machine Translation

(1) , (1) , (1) , (1)
1

Abstract

A confidence measure is able to estimate the reliability of an hypothesis provided by a machine translation system. The problem of confidence measure can be seen as a process of testing : we want to decide whether the most probable sequence of words provided by the machine translation system is correct or not. In the following we describe several original word-level confidence measures for machine translation, based on mutual information, n-gram language model and lexical features language model. We evaluate how well they perform individually or together, and show that using a combination of confidence measures based on mutual information yields a classification error rate as low as 25.1\% with an F-measure of 0.708.
Fichier principal
Vignette du fichier
confidence_measures-article.pdf (288.37 Ko) Télécharger le fichier
Vignette du fichier
slides.pdf (529.09 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Format : Other

Dates and versions

inria-00333843 , version 1 (30-01-2009)

Identifiers

  • HAL Id : inria-00333843 , version 1
  • ARXIV : 0902.1033

Cite

Sylvain Raybaud, Caroline Lavecchia, David Langlois, Kamel Smaïli. New Confidence Measures for Statistical Machine Translation. International Conference On Agents and Artificial Intelligence - ICAART 09, Jan 2009, Porto, Portugal. ⟨inria-00333843⟩
82 View
198 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More