Building Word-Aligned Bilingual Corpus for Statistical Myanmar-English Translation

In recent years statistical word alignmentmodels have been widely used for various NaturalLanguage Processing (NLP) problems. In this paperwe describe our work in constructing an alignedEnglish-Myanmar parallel corpus. Corpora are notavailable for Myanmar language and our work indeveloping parallel corpus will also hopefully bevery useful in many natural languageapplications. Word alignment plays a crucial role instatistical machine translation, since word-alignedcorpora have been found to be an excellent source oftranslation-related knowledge. If there were errors inalignment, this will cause subsequence failure NLPprocesses. The alignments produced when thetraining on word-aligned data are dramaticallybetter than when training on sentence-aligned data.The main purpose of this system is to provide as partof translation machine in Myanmar-English machinetranslation. The proposed system is combination ofcorpus based approach and dictionary lookupapproach. The corpus based approach is based onthe first three IBM models