Question: Let consider B a corpus which only contains one single bitstring: 110111001011101111000 (a) (1.5 points) Calculate the following bigram probabilities from the corpus B using

Let consider B a corpus which only contains one single bitstring: 110111001011101111000 (a) (1.5 points) Calculate the following bigram probabilities from the corpus B using MLE (Maximum Likelihood Estimation). Answer with a ratio p/q, not a floating point number. P(01) and P(00) (b) (1.5 points) Assume a bigram language model created from corpus B. For each of the following bitstrings, decide if it is more probable that x1 resp x2 is 0 or 1 . 1010101x1 and 0101010x2 (c) (2 points) Calculate the following bigram probabilities from the corpus B using Laplacian smooting. Answer with a ratio p/q, not a floating point number. P(01),P(11),P(00) and P(10) (d) (1 point) What is the point of using a Laplace smoothing in this case
Step by Step Solution
There are 3 Steps involved in it
Get step-by-step solutions from verified subject matter experts
