TnT - Statistical Part-of-Speech Tagging

Thorsten Brants · Americanae (AECID Library) · 2000

TnT - Trigrams n Tags - is a very efficient statisticalpart-of-speech tagger that is trainable on different languages andvirtually any tagset. The tagger comes with two pre-compiled languagemodels, one for German using STTS, and two for Engli sh using the PennTreebank tagset and the Susanne tagset.TnT can be adapted to any language model that delimit words with white space.

Read the paper · More papers on PaperTik