Abbott, JZMartinus, Laura JB2019-02-242019-02-242018-12Abbott, J.Z. and Martinus, L.J.B. 2018. Towards neural machine translation for African languages. NIPS 2018 Workshop on Machine Learning for the Developing World, December 2018, Montreal, Canadahttps://arxiv.org/html/1812.10398http://hdl.handle.net/10204/10726Paper presented at the NIPS 2018 Workshop on Machine Learning for the Developing World, December 2018, Montreal, CanadaGiven that South African education is in crisis, strategies for improvement and sustainability of high-quality, up-to-date education must be explored. In the migration of education online, inclusion of machine translation for low-resourced local languages becomes necessary. This paper aims to spur the use of current neural machine translation (NMT) techniques for low-resourced local languages. The paper demonstrates state-of-the-art performance on English-to-Setswana translation using the Autshumato dataset. The use of the Transformer architecture beat previous techniques by 5.33 BLEU points. This demonstrates the promise of using current NMT techniques for African languages.enNeural machine translationNMTTransformer architectureConference PresentationAbbott, J., & Martinus, L. J. (2018). Towards neural machine translation for African languages. http://hdl.handle.net/10204/10726Abbott, JZ, and Laura JB Martinus. "Towards neural machine translation for African languages." (2018): http://hdl.handle.net/10204/10726Abbott J, Martinus LJ, Towards neural machine translation for African languages; 2018. http://hdl.handle.net/10204/10726 .TY - Conference Presentation AU - Abbott, JZ AU - Martinus, Laura JB AB - Given that South African education is in crisis, strategies for improvement and sustainability of high-quality, up-to-date education must be explored. In the migration of education online, inclusion of machine translation for low-resourced local languages becomes necessary. This paper aims to spur the use of current neural machine translation (NMT) techniques for low-resourced local languages. The paper demonstrates state-of-the-art performance on English-to-Setswana translation using the Autshumato dataset. The use of the Transformer architecture beat previous techniques by 5.33 BLEU points. This demonstrates the promise of using current NMT techniques for African languages. DA - 2018-12 DB - ResearchSpace DP - CSIR KW - Neural machine translation KW - NMT KW - Transformer architecture LK - https://researchspace.csir.co.za PY - 2018 T1 - Towards neural machine translation for African languages TI - Towards neural machine translation for African languages UR - http://hdl.handle.net/10204/10726 ER -