A neural machine translation approach for translating Malay parliament Hansard to English text

Parliament Hansard is one of the most precious texts made available to the public. In Malaysia, the parliament Hansard records the debate and discussions in Malay language. Topic modelling, sentiment analysis, relation extractions, trend prediction or temporal analyses are frequently applied on parliament Hansard to discover interesting patterns. However, most of the matured tools for such processing tasks work on English text. As such, before the Malaysian parliament Hansard can be further processed, it is essential to translate the Malay text into English. Several machine translation approaches have been surveyed in this paper. From the literature review, neural machine translation, particularly the Transformer Model has been proven to provide promising results in translating different languages. In this paper, we present our implementation of neural machine translation for Malay to English text. The experimental design shows that with a good set of parallel corpus and minimal fine-tuning, neural MT can achieve as high as 35.42 in BLEU score.