Evaluation of English-Myanmar Syntactic Reordering System by Using Reordering Metrics

Abstract

Because different languages employ different word order in their syntax, one requirement of machine translation (MT) system is to get the target word in the right order. While phrasebased machine translation system do very well at short-distance reordering, long-distance reordering seems to be a challenging task. One way of overcoming this challenge is to use linguistic information and reorder the input sentence so that the word order is consistent with what the target language might expect. In this paper, long-distance reordering is solved by using the linguistic information such as part-of – speech tags and syntactic function tags. Moreover, the quality of word order is also measured by using reordering metrics. This reordering system is implemented for using as a component in English-Myanmar Translation system.