Agglutinative languages encode a substantial part of grammatical relations through ordered affix sequences attached to lexical stems. Encoders in the BERT family can learn some of these regularities implicitly from large corpora, but conventional self-attention does not treat lemma, stem/root, part...

This paper presents a part-of-speech tagging system for the Uzbek language based on the Conditional Random Fields (CRF) approach. Using a manually annotated corpus and a set of language-specific morphological and contextual features, the model was trained and evaluated through 5-fold...