POS-tagging-with-Vterbi-algorithm
Using HMM for pos tagging task
Here I use Viterbi algorithm to do pos tagging task.
http://universaldependencies.org/#en).
Dataset: dataset in English from Universal Dependencies (http://universaldependencies.org/en/pos/all.html) to see all possible tagsin the corpus.
Tags: Refer this link (- train.counts: Frequency counts from the training file. It has two types of counts: (a) Emission counts, in the format WORDTAG (b) n-gram counts (where N is 1, 2 or 3), in the format <tag 1> ...
- test.words: Simplified version of the original test file, with one word per line and blank lines separating sentence.
- test.tags: Simplified version of the original test file, with word and the true tag on each line and blank lines separating sentences.
Also, plz refer README.pdf for specific instructions.