Generate classification probabilities using a Naive Bayes classifier

scikit-learn has naive bayes classifier which can produce probabilities of classification. It would be good to try this out. If we get reasonable accuracy and some probabilities, this could be very useful as an easy to create probabilities for use in downstream processing.

https://scikit-learn.org/stable/modules/generated/sklearn.naive_bayes.MultinomialNB.html#sklearn.naive_bayes.MultinomialNB

Start with the linked scikit-learn MultinomialNB documentation and the repository's existing classification workflow. Determine how classifier probabilities and accuracy should be evaluated, and how useful probabilities would be exposed for downstream processing. Done means an agreed implementation produces reasonable classification accuracy and usable probabilities.

Written by the indexing model from the issue text.