Using NLTK, train a HMM model and return POS tags by Viterbi algorithm

Part 1: TRAINING A HIDDEN MARKOV MODEL Create and train two models - an Emission Model and a Transition Model

• Clean and structure data
• Use ConditionalProbDist with a LidstoneProbDist estimator +0.01 added to the sample count for each bin.

Part 2: IMPLEMENTING THE VITERBI ALGORITHM (for POS tagging) Implement the Viterbi algorithm

• 2 data structures for the algorithm: the viterbi data structure and the backpointer data structure
• use costs (negative log probabilities)
• recursion and termination step of algorithm
• test the algorithm