International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 129 - Number 10 |
Year of Publication: 2015 |
Authors: Safiriyu I. Eludiora, Abayomi O. Agbeyangi, O.I. Fatusin |
10.5120/ijca2015907007 |
Safiriyu I. Eludiora, Abayomi O. Agbeyangi, O.I. Fatusin . Development of English to Yorùbá Machine Translation System for Yorùbá Verbs’ Tone Changing. International Journal of Computer Applications. 129, 10 ( November 2015), 12-17. DOI=10.5120/ijca2015907007
In translating English sentences (text) to Yorùbá sentences (text), some Yorùbá verbs change tone from the bilingual dictionary low-tone to mid-tone when they are translated to Yorùbá. They are called tone change verbs. These tone change verbs do pose some challenges in English to Yorùbá machine translation. Most of the time it changes the meaning of the sentence. These changes usually depend on the positions of the nouns and pronouns in the sentence. The verbs in this category were collected from different Yorùbá sentences that contain tone change verbs. Some re-write rules were designed for the two languages. The re-write rules were tested using JFLAP. Apart from re-write rules, there are other grammatical rules considered and the rules affected the Yorùbá translations. The software was designed using unified modelling language (UML). The Rule-based approach was used for the translation. Python programming language was used for the software development. The python has natural language tool kits that are used for the sentence parsing. The system accept English sentence then discover the pattern for the sentence. The system was implemented and tested for twenty tone change verbs within the home domain. The two languages are subject verb object (SVO) sentence structure with some differences. The results show that the MT system can translate these tone change verbs. The system is efficient in its response time. IFEMT2 system was developed to handle the translation of tone changing verbs. IFEMT2 was used to compare with the Google translator. IFEMT2 system performs better than the Google translator. The Google translator could not adequately translate these tone change verbs.