An workflow intended to demonstrate how you can extract patterns of parts of speech (for example, verb followed by adverb followed by noun) and count/visualize those patterns. Caveats: * No usual text pre-processing steps are applied here to clean up the data * Workflow assumes sequences of size 3 * Sequences extend across sentences, which may not make sense * Sequences are counted across the entire corpus, instead of by document The workflow uses the first 5 rows of the IMDB movie review dataset.
This is a companion discussion topic for the original entry at https://kni.me/w/Vp3wxDZLXi19gFMY