English to Yoruba short message service speech and text translator for android phones

Autor: Afeez Adeshina Oke, Hafsat Omolola Olawuyi, Abdulkarim A. Oloyede, Christiana Oluwakemi Abikoye, Roseline Oluwaseun Ogundokun, Akinbowale Nathaniel Babatunde
Rok vydání: 2021
Předmět:
Zdroj: International Journal of Speech Technology. 24:979-991
ISSN: 1572-8110
1381-2416
Popis: Machine language translation (MLT) has acquired a substantial quantity of investigation consideration in Europe and Asia, but works on African languages, especially the Yoruba language, are rare. There is, however, a communication barrier between the people who solely speak Yoruba and foreign visitors and places where the language is not being said or understood. There is a necessity to focus on the glitches of English to Yoruba machine translations to bridge the communication gap. This study, therefore, implemented the rule-based machine translation technique. The method used involved formulating 20 computational rules for the translation process. Offline Android-based English to Yoruba short message service application was developed using the Android Studio IDE and the some of the available inbuilt android plugins alongside a locally produced bilingual dictionary. This paper also examined translations done by the developed mobile app contrary to human translation to explore why machine translation systems still give faults in interpreting natural human language. A questionnaire was designed for the translations; they were distributed to one hundred individuals. Eighty-one participants gave their feedbacks. The responses were dispensed to statistical examination using SPSS. Discoveries showed that human translation handled best-concerning precision and eloquence, which are knowledgeable by the quality and the quantity of the computational rules formulated.
Databáze: OpenAIRE
Externí odkaz: