A Rule-based Approach to English-Okun Prepositional Phrase Machine Translation

Abstract

Okun culture is gradually going into extinction because the language is greatly dominated by English language. Therefore, this research  developed a rule-based English-Okun machine translator using prepositional phrase. The dataset used were obtained from locally spoken  words and the correctness of the re-write rules was tested using Java Formal Language Automata Package. Human judgment  was employed in evaluating the system based on orthography accuracy. Results show that the developed translator is 95 percent  accurate while experimental subject respondents have an accuracy of 58 percent. The developed machine translator’s performance is  closer to the Experts’ but higher than that of the experimental subject respondents, this shows that indigenous people are not good at  writing Okun language and find it hard to separate the dialect from Yoruba language. Hence, research in EnglishOkun machine  translation should be intensified in order to prevent the language from going into extinction and future research should consider  extending the full form bilingual lexicon of Okun language to achieve improved accuracy.