I am an enthusiastic proponent of using tidy data principles for dealing with text data. This kind of approach offers a fluent and flexible option not just for exploratory data analysis, but also for machine learning for text, including both unsupervised machine learning and supervised machine learning. I haven’t written much about supervised machine learning for text, i.e. predictive modeling, using tidy data principles, so let’s walk through an example workflow for this a text classification …