Thai to Isarn dialect machine translation using rule-based and example-based
Paweena Unlee, Pusadee Seresangtakul · 2016
This paper presented a Thai to Isarn dialect machine translation, named ThaiIsarn - MT, using rule-based and example based. To develop the system, a Thai - Isarn dialect dictionary and linguistics rules database were constructed. The dictionary contained 8,050 Thai words. Moreover, an example bilingual corpus was constructed. Since Thai is non-segmented language, input sentence was segmented to sequence of word using longest matching algorithm. The Thai word was looked for the corresponding Isarn word from the dictionary. For word that may has several meanings, we used the rule based and example based to select the most suitable meaning. In order to evaluate the performance of the purposed system, 854 Thai sentences were translated by the system. The result shows that the translation accuracy is 82.44%.