DFKI-LT - A Linguistic Evaluation of Rule-Based, Phrase-Based, and Neural MT Engines
A Linguistic Evaluation of Rule-Based, Phrase-Based, and Neural MT Engines
4 The Prague Bulletin of Mathematical Linguistics volume 108 number 1,
In this paper, we report an analysis of the strengths and weaknesses of several Machine Translation (MT) engines implementing the three most widely used paradigms. The analysis is based on a manually built test suite that comprises a large range of linguistic phenomena. Two main observations are on the one hand the striking improvement of an commercial online system when turning from a phrase-based to a neural engine and on the other hand that the successful translations of neural MT systems sometimes bear resemblance with the translations of a rule-based MT system.
Files: BibTeX, pralin-2017-0017.pdf, [The Prague Bulletin of Mathematical Linguistics] A Linguistic Evaluation of Rule-Based Phrase-Based and Neural MT Engines.pdf