Home Linguistics & Semiotics Preprocessing for English-to-Arabic Statistical Machine Translation
Chapter
Licensed
Unlicensed Requires Authentication

Preprocessing for English-to-Arabic Statistical Machine Translation

  • Rabih Zbib and Ibrahim Badr
View more publications by John Benjamins Publishing Company

Abstract

Most research in Arabic Statistical Machine Translation (SMT) has focused on translating from Arabic into English and other languages. Translation to Arabic has drawn very little attention so far, despite being an important, as well as technically challenging task. This chapter describes the application of two preprocessing techniques to English-to-Arabic SMT: Morphological Segmentation and Syntactic Reordering. It shows how these techniques can be adapted to apply to translation into Arabic, providing significant improvements to a phrase-based system.

Abstract

Most research in Arabic Statistical Machine Translation (SMT) has focused on translating from Arabic into English and other languages. Translation to Arabic has drawn very little attention so far, despite being an important, as well as technically challenging task. This chapter describes the application of two preprocessing techniques to English-to-Arabic SMT: Morphological Segmentation and Syntactic Reordering. It shows how these techniques can be adapted to apply to translation into Arabic, providing significant improvements to a phrase-based system.

Downloaded on 17.9.2025 from https://www.degruyterbrill.com/document/doi/10.1075/nlp.9.06zbi/html
Scroll to top button