A framework for implementation of Thai to English machine translation systems

Abstract

Machine translation (MT) is a sub-field of computational linguistics that investigates the use of computer software to translate text or speech from one language to another. It can help people to communicate with others with different natural languages. Machine Translation (MT) between Thai and English languages bas been an ongoing and challenging research topic. Most research has been done on English to Thai machine translation. but not the other way around. Developing a machine translation system which can correctly translate any sentence from one language to another, seems to be a formidable and challenging task due to the big differences between syntax and semantics of the two languages. This suggests that such a system should be implemented incrementally. Starting with a system that can translate simple phrases and sentences, it can then be expanded to include grammar rles and capabilities to translate new sentences which are more complex and have not previously been successfully translated. This research proposes a new framework to develop Thai to English machine translation systems incrementally. It is based on the interlingua approach, where input sentences are translated into internediate representations before they are converted into equivalent sentences in the other language. Lexical Functional Grammar (LFG) - or an LFG tree structure - is used as an intermediate represe entation in this research. The framework has been used to implement a prototype system that can translate about 200 Thai phrases and sentences into English phrases and sentences, resulting in a satisfactory quality of the translation output.