nlp subject Exercise 7 audit results fail to follow task instructions

[jarmo-seljamaa]

nlp

We struggled to get the results as described in the audit of Exercise 7. After a lot of trial and error we discovered that we get the exact audit results ONLY when we ignore the task instructions.

The instructions say:

Steps:

1. Preprocess the data using the function implemented earlier. Then, use CountVectorizer from scikit-learn with max_features=500 to compute the word count of the tweets. The output is a sparse matrix.

We could get the expected audit results for Question 1, 2 and 3 when we skip preprocessing completely and feed the raw data from the given file into the CountVectorizer. Most probably this is not the intended solution. Please re-check the audit.

[nprimo]

Hi @jarmo-seljamaa, thank you for the feedback! The subject and audit are going to be updated in the following days to resolve this issue.