A study on customized neural machine translation exemplified by IB MYP Summative Assessment criteria

Neural Machine Translation (NMT) is a recent research area in translation industry. Various studies are being conducted on the use of NMT in different fields of study. One of these fields is education. Education is an expanding area with high volumes of data. In this context, the International Baccalaureate (IB) is one of the internationally accepted educational programmes. Great volumes of translation needs are required to be met continuously in the IB programme. Therefore, customized NMT technology is used to provide a solution to high and constant translation needs of the IB in this thesis. The results obtained from the customized NMT system are compared with those of Statistical Machine Translation (SMT). The findings are analyzed within the scope of Multidimensional Quality Metrics (MQM) and BLEU Score metric system. Finally, a conclusion is drawn in line with the comparisons and analyses.