CountVectorizer and TF/IDF as NLP pre-processing #62
Description
To deal with textual data at a basic level could offer a bag of words processing (instead to one-hot). TF/IDF might improve the result but is optional. It creates columns per type and values for token counts.