Small Typos in C2 W2 Assignment

Course Q&A Natural Language Processing NLP with Probabilistic Models
1

Hey Guys,
I observed the following discrepancies in the C2 W2 Assignment. Please take them into account for future revisions.

Part 0: Data Sources

The training set will be used to create the emission, transmission transition and tag counts.

… training and test corpus and will appear in the emission, transmission transition and tag data structures.

  • In the image in this section as well, it is mentioned transmission_counts, whereas, it is supposed to be transition_counts.

Exercise 01

You can get a more complete description at Penn Treebank II tag set.

The above URL doesn’t work any more.

Exercise 03

Now you will create the B transition emission matrix which computes the emission probability.

Part 3.2 Viterbi Forward

Compute the probability that the tag of the second work word (‘tracks’) is a verb, 3rd person singular present (VBZ).

Exercise 06

  • In the image above the UNQ C6, the figure depicts the best_paths matrix.
  • In that, we have mentioned the value for (--s--, 'tracks') as 32.
  • Now, this matrix stores the indices of the most likely tags for the previous words, however, in the figure, no tag has been indicated with the index 32.
  • It would be great if this index can be changed to any other index mentioned in the figure, to avoid any confusion whatsoever.

Cheers,
Elemento

2

Hi Elemento,

Great catches! I noticed that @Mubsi has opened a github issue to address them.