Sentence alignment for english urdu language pair

Abstract

Sentence aligned parallel text is an important resource in statistical machine translation; therefore, Sentence alignment is a crucial part of machine translation.The sentence alignment task comprises of recognizing the correspondence between words sentences and paragraphs of the source and target languages. Different researchers proposed several sentence alignment algorithms for aligning sentences of the source and target language. In this paper, we have explored sentence alignment algorithms based on character length, word length and lexical matching and carry out performance analysis of Gale and Church Algorithm on English and low resource language Urdu.