The Evaluations of Thai Poetry Translator to English with Prosody Keeping

Poetry Machine Translator is problematic area challenging. It is very important that output of poetry translation should still be poetry. Thai Poetry is composed with sets of syllables. Those rhymes, existing from stanzas, lines and the text in paragraph of the poetry, may not represent the complete syntax. Evaluation of translation will show the quality of translator but Evaluation is still a research topic in itself. This researchwill focus on the Thai poetry type “Klonn-Pad” and aim to translate into English keeping terms of prosody. The results of translation between Forward and Backward translation will be compared by BLEU (Bilingual Evaluation Understudy) metric and METEOR (Metric for Evaluation of Translation with Explicit Ordering). Prosody keeping will be check by Error percentage of prosody rule. The prosody error checking separate to Check line length (Number of syllable) and Check Rhyme (Relation of phonetic). The original (Thai) language L1 translates to the target language L2. The result of Forward is L2out and the result of Backward is L1out. The BLEU score of Forward (L2out) equaled 0.751, Backward (L1out) equaled 0.885. The METEOR score of Forward (L2out) Precision equaled 0.863, Recall equaled 0.791 and F-Measure equaled 0.840. For Backward (L1out) Precision equaled 0.850, Recall equaled 0.863 and F-Measure equaled 0.856. Check line length correctness of Forward (L2out) equaled 0.875, Backward (L1out) equaled 1. Check Rhyme correctness of Forward (L2out) equaled 0.857, Backward (L1out) equaled 0.643. Based on this study, it can be concluded that the Thai poetry machine translators with Poetry’s dictionary and Tuning Module can provide the good result. Thus, it is necessary and reasonable that Synonyms Dictionary of Thai Poetry should be develop and improve.