After assessing the situation, it was clear several hurdles existed that would need to be overcome in order to successfully translate the high volume of materials and meet the critical deadlines of each cycle.
One difficulty was that the client-specific corpora needed to be accurately and quickly translated. Corpora can include anything from company logos to product names and lingo that are used solely by the client. Other in-market engines are incapable of deciphering not only the corpora, but the more complicated, non-traditional languages themselves, like Vietnamese.
To meet the turnaround requirements and ensure the quality of the output across languages and corpora, it was determined a customer-specific NMT engine would need to be created.
ULG’s dedicated NMT team gathered all necessary industry and client specific content before completing the following steps: data cleaning and preparation, training and multiple levels of testing. These phases were finished before a customized engine could be released for utilization within the translation workflow.
