About this document
TnT: Efficient POS Tagging Model by tejshahi1 is a document available to read on EtoBox.
TnT is an efficient statistical part-of-speech tagger based on Markov models that performs as well or better than other current approaches. It uses second-order Markov models to calculate tag probabilities based on previous tags and words. It smooths trigram probabilities using linear interpolation and handles unknown words by analyzing suffixes to determine probable tags. Evaluations on two corpora show TnT achieves high accuracy while also being fast to train and use for tagging.
- Author
- tejshahi1
- Language
- EN