A Survey of Machine Translation Tasks on Nigerian Languages

Proceedings of the 13th Conference on Language Resources and Evaluation (LREC 2022), pages 6480–6486 Marseille, 20-25 June 2022 © European Language Resources Association (ELRA), licensed under CC-BY-NC-4.0 6480 A Survey of Machine Translation Tasks on Nigerian Languages Ebelechukwu Nwafor1, Anietie Andy2 1Department of Computing Sciences, 2 Penn Medicine. 1Villanova University, Villanova, PA, 19085, 2University of Pennsylvania, Philadephia, PA, 19104. Email: ebelechukwu.nwafor@villanova.edu, anietie.andy@pennmedicine.upenn.edu Abstract Machine translation is an active area of research that has received a significant amount of attention over the past decade. With the advent of deep learning models, the translation of several languages has been performed with high accuracy and precision. In spite of the development in machine translation techniques, there is very limited work focused on translating low-resource African languages, particularly Nigerian languages. Nigeria is one of the most populous countries in Africa with diverse language and ethnic groups. In this paper, we survey the current state of the art of machine translation research on Nigerian languages with a major emphasis on neural machine translation techniques. We outline the limitations of research in machine translation on Nigerian languages and propose future directions in increasing research and participation. Keywords: Machine translation, low-resource languages 1. Introduction Machine translation is an important and very beneficial computing task that provides an automated mechanism for communicating across languages. In recent years, there has been an increase in available tools and techniques for ma- chine translation such as statistical (Koehn et al., 2007b) and neural machine translation (Bahdanau et al., 2016; Lu- ong et al., 2015) techniques. Recently, neural machine translation has become the standard means for language translation tasks with advanced techniques such as encoder- decoder models (Vaswani et al., 2017) which have proven to produce remarkable state of the art results. With the advancements in neural machine translation, there has been a lot of work in the literature which focuses on machine translation task of high-resource languages (Wu et al., 2016; Koehn and others, 2005) such as from English to French (Meyer et al., 2013; Brown et al., 1988; He et al., 2016) or English to German (Cheng et al., 2018). How- ever, there are very few literature on machine translation of African languages (Nekoto et al., 2020b; Abbott and Marti- nus, 2018) and even fewer works centered on Nigerian lan- guages. This might be attributed to socio-economic issues such as the high cost of training language models and the effort involved in curating high-quality datasets. Nigeria is one of the most populous country in Africa (Akinyemi and Isiugo-abanihe, 2014) with a population of over 150 mil- lion inhabitants and over 500 languages 1. According to the United Nations (noa, ), Nigeria is the 7th largest world population, and it is estimated that the population of Nige- ria will surpass that of the United States and become the third largest nation before 2050. In addition to being one of the most populous country in Africa with immense human resources, Nigeria is poised for business growth with an in- crease in emerging …

Content truncated to 500 words