c. Suppose that an NLP Engine wants to tag the sequence,"change is good" using 3 possible tag A, B and C. The engine has
Posted: Sun Jul 10, 2022 10:51 am
c. Suppose that an NLP Engine wants to tag the sequence,"change is good" using 3 possible tag A, B and C. The engine has the following probabilities information from training data: P(change(A)=1/5, P(change B)=1/5, P(change|C)=4/5 P(is|A)=2/5, P(is|B)=0, P(is|C)=0, P(good|A)=1/10, P(good|B)=1/3, P(good|C)=1/3 P(A/A)=1/5,P(BIA)=1/5,P(CIA)=3/5 P(A/B)=1/5,P(B|B)=8/10,P(CIB)=0 P(A/C) 2/3,P(BIC)=1/3,P(CIC)=0 Assume that all the tags have the same probabilities at the beginning of the sentence (and that is 1/3 each). Find out the best tag sequence using the Viterbi algorithm along with value at each vertex. [5marks]