Centrum für Informations- und Sprachverarbeitung
print


Breadcrumb Navigation


Content

Fine-grained morphological tagging for 17 languages

18.03.2015

A new version of the morphological tagger MarMoT is available. The release includes pretrained models for 17 languages: Arabic, Basque, Bulgarian, Czech, English, Estonian, Farsi, French, German, Modern Hebrew, Hungarian, Korean, Polish, Romanian, Serbian, Slovak, Slovenian and Swedish. The models were trained on Multext-East V4 and SPMRL 2013.

MarMot was developed by Thomas Müller and is avalable as open source.


Service