Semantic Machine Translation Engine – solving major NLP problems empowered by semantics
LYLO does not research generic AI approaches for NLP. Instead, we use all NLP legacy (e.g. POS and Dependency parsing from Stanford, phrasal translation from Google, FrameNet from Berkeley, WordNet from Princeton) combined with proprietary grammar induction and semantic pattern building mechanisms. Our approach brings English - German pilot translation to best in class within 21 months. Customers - Google, Apple or as API for end users.