Statistical translation of text and speech

First results with the RWTH system

C. Tillmann, Stephan Vogel, H. Ney, H. Sawaf

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

In this paper, we describe a first version of a system for statistical translation and present experimental results. The statistical translation approach uses two types of information: a translation model and a language model. The language model used is a standard bigram model. The translation model is decomposed into lexical and alignment models. After presenting the details of the alignment model, we describe the search problem and present a dynamic programming-based solution for the special case of monotone alignments. So far, the system has been tested on two limited-domain tasks for which a bilingual corpus is available: the EuTrans traveller task (Spanish-English, 500-word vocabulary) and the Verbmobil task (German-English, 3000-word vocabulary). We present experimental results on these tasks. In addition to the translation of text input, we also address the problem of speech translation and suitable integration of the acoustic recognition process and the translation process.

Original languageEnglish
Pages (from-to)43-74
Number of pages32
JournalMachine Translation
Volume15
Issue number1-2
DOIs
Publication statusPublished - 1 Dec 2000
Externally publishedYes

Fingerprint

vocabulary
language
Dynamic programming
acoustics
programming
Acoustics
present
Alignment
Vocabulary
English Words
Language Model
Travellers
Translation Process
Programming

Keywords

  • Bilingual alignment
  • Dynamic programming
  • Hidden Markov models
  • Learning from bilingual corpora
  • Statistical machine translation

ASJC Scopus subject areas

  • Hardware and Architecture
  • Software

Cite this

Statistical translation of text and speech : First results with the RWTH system. / Tillmann, C.; Vogel, Stephan; Ney, H.; Sawaf, H.

In: Machine Translation, Vol. 15, No. 1-2, 01.12.2000, p. 43-74.

Research output: Contribution to journalArticle

Tillmann, C. ; Vogel, Stephan ; Ney, H. ; Sawaf, H. / Statistical translation of text and speech : First results with the RWTH system. In: Machine Translation. 2000 ; Vol. 15, No. 1-2. pp. 43-74.
@article{0adbadec3c7d487797977f96d9955d62,
title = "Statistical translation of text and speech: First results with the RWTH system",
abstract = "In this paper, we describe a first version of a system for statistical translation and present experimental results. The statistical translation approach uses two types of information: a translation model and a language model. The language model used is a standard bigram model. The translation model is decomposed into lexical and alignment models. After presenting the details of the alignment model, we describe the search problem and present a dynamic programming-based solution for the special case of monotone alignments. So far, the system has been tested on two limited-domain tasks for which a bilingual corpus is available: the EuTrans traveller task (Spanish-English, 500-word vocabulary) and the Verbmobil task (German-English, 3000-word vocabulary). We present experimental results on these tasks. In addition to the translation of text input, we also address the problem of speech translation and suitable integration of the acoustic recognition process and the translation process.",
keywords = "Bilingual alignment, Dynamic programming, Hidden Markov models, Learning from bilingual corpora, Statistical machine translation",
author = "C. Tillmann and Stephan Vogel and H. Ney and H. Sawaf",
year = "2000",
month = "12",
day = "1",
doi = "10.1023/A:1011168216856",
language = "English",
volume = "15",
pages = "43--74",
journal = "Machine Translation",
issn = "0922-6567",
publisher = "Springer Netherlands",
number = "1-2",

}

TY - JOUR

T1 - Statistical translation of text and speech

T2 - First results with the RWTH system

AU - Tillmann, C.

AU - Vogel, Stephan

AU - Ney, H.

AU - Sawaf, H.

PY - 2000/12/1

Y1 - 2000/12/1

N2 - In this paper, we describe a first version of a system for statistical translation and present experimental results. The statistical translation approach uses two types of information: a translation model and a language model. The language model used is a standard bigram model. The translation model is decomposed into lexical and alignment models. After presenting the details of the alignment model, we describe the search problem and present a dynamic programming-based solution for the special case of monotone alignments. So far, the system has been tested on two limited-domain tasks for which a bilingual corpus is available: the EuTrans traveller task (Spanish-English, 500-word vocabulary) and the Verbmobil task (German-English, 3000-word vocabulary). We present experimental results on these tasks. In addition to the translation of text input, we also address the problem of speech translation and suitable integration of the acoustic recognition process and the translation process.

AB - In this paper, we describe a first version of a system for statistical translation and present experimental results. The statistical translation approach uses two types of information: a translation model and a language model. The language model used is a standard bigram model. The translation model is decomposed into lexical and alignment models. After presenting the details of the alignment model, we describe the search problem and present a dynamic programming-based solution for the special case of monotone alignments. So far, the system has been tested on two limited-domain tasks for which a bilingual corpus is available: the EuTrans traveller task (Spanish-English, 500-word vocabulary) and the Verbmobil task (German-English, 3000-word vocabulary). We present experimental results on these tasks. In addition to the translation of text input, we also address the problem of speech translation and suitable integration of the acoustic recognition process and the translation process.

KW - Bilingual alignment

KW - Dynamic programming

KW - Hidden Markov models

KW - Learning from bilingual corpora

KW - Statistical machine translation

UR - http://www.scopus.com/inward/record.url?scp=0348181952&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0348181952&partnerID=8YFLogxK

U2 - 10.1023/A:1011168216856

DO - 10.1023/A:1011168216856

M3 - Article

VL - 15

SP - 43

EP - 74

JO - Machine Translation

JF - Machine Translation

SN - 0922-6567

IS - 1-2

ER -