A COMPUTATIONAL MODEL OF ENGLISH TO YORUBA NOUN-PHRASES TRANSLATION SYSTEM
Oluwatoyin Bunmi Abiola · Futa Journal of Research in Sciences · 2014
The field of natural language processing enables machines to read and understand the languages human being speaks. Developing a computational model for English language and Yoruba language noun-phrases involve a profound understanding of the syntactic and grammatical features of the two languages as well as their vocabularies since they are not related syntactically and grammatically. Twenty nine rules were formulated for the noun phrase translations which were specified using the Context Free Grammar (CFG).We then modeled and recognized the grammar of the language using the Finite State Automata (FSA)whose operations was based on the first set techniques. The first sets techniques allow the parser to choose which production rule to apply based on the first input word of an input phrase. We also developed a bilingual lexicon which is made up of words in English language with their corresponding Yoruba counterparts and their equivalent part of speech. The model was implemented using PhP programming language and MySQL and was tested on one hundred and sixty randomly selected noun-phrases from daily news, and gives accuracy of 91% which is quite encouraging. The system if fully developed will go a long way in preventing the extinction threat of the Yoruba language. Keywords: Natural Language Processing, English, Yoruba, Noun-Phrases, Translation System, Context- Free Grammar and Finite State Automata.