Skip to main content


Neural machine translation : Tando


Following the significant advances made by neural paradigms in the area of machine translation, one of the main challenges existing today is to increase consistency on the paragraph or document level of the systems. Most of today’s systems translate each sentence in isolation without taking the general context in which the sentence appears into consideration. This leads to consistency problems in the discourse. So this project aims to conduct research into and develop architectures and algorithms that will take the context of the sentences into consideration and thus significantly improve translation consistency and overall quality.


Another similar projects