This paper describes efforts towards the development of an Arabic to Italian SMT system for the news domain. Since only very little parallel data are available for this language pair, we investigated both the exploitation of comparable corpora and pivot translation. Experimental evaluation was conducted on a new benchmark developed by extending two Arabic-to-English NIST evaluation sets with Italian and French translations, produced from the source language by experts. Preliminary results show potentials of both approaches with respect to performance achieved by a popular state-of-the-art Web-based translation service.

Bootstrapping Arabic-Italian SMT through Comparable Texts and Pivot Translation

Cettolo, Mauro;Bertoldi, Nicola;Federico, Marcello
2011-01-01

Abstract

This paper describes efforts towards the development of an Arabic to Italian SMT system for the news domain. Since only very little parallel data are available for this language pair, we investigated both the exploitation of comparable corpora and pivot translation. Experimental evaluation was conducted on a new benchmark developed by extending two Arabic-to-English NIST evaluation sets with Italian and French translations, produced from the source language by experts. Preliminary results show potentials of both approaches with respect to performance achieved by a popular state-of-the-art Web-based translation service.
File in questo prodotto:
File Dimensione Formato  
EAMT-2011-Cettolo-etAl.pdf

accesso aperto

Tipologia: Documento in Post-print
Licenza: PUBBLICO - Pubblico con Copyright
Dimensione 353.64 kB
Formato Adobe PDF
353.64 kB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11582/31564
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
social impact