Part-of-Speech Tagging with Two Sequential Transducers

Andre Kempe

We present a method of constructing and using a cascade consisting of a left- and a right-sequential finite-state transducer (FST), T1 and T2, for part-of-speech (POS) disambiguati…