Jordan Boyd-Graber
There’s an error defining the transition matrix. It should be theta_{i,j} = p(z_n = j | z_{n-1} = 1), not beta. .(tagsToTranslate)HMM(t)Hidden Markov Model(t)NLTK(t)Computational Linguistics (Job Title)
Similar Posts
2 thoughts on “Computational Linguistics I: Part of Speech Tagging”
Comments are closed.
thanks for this tutorial .. can you please tell me how to use forward backward algorithm in POS training process .. ?
This video does let me have a quick glance at the POS tagging, but it seems that the description of Viterbi Algorithm in the slides has something wrong —— in (4), the transition term theta_{j,k} should be removed.
based on wiki, my personal calculation and a lecture note
http://ttic.uchicago.edu/~dmcallester/ttic101-06/lectures/hmm/HMMs.pdf