POS tagger for Hidden Markov Model using Trigrams
Though we have many POS taggers which implement Viterbi algorithm. But the challenge is to get a better accuracy and assigning accurate tag to a word. I propose a model which will implement Viterbi algorithm over trigrams with the help of a smoothing technique.
With the given corpora and the results of the proposed model, it can be concluded that the accuracy for a language model would increase with the increase in the n-grams. Though the project uses uniform distribution for identifying the unseen words, the experiments with other techniques such as applying morphological rules is also giving a close accuracy results.