The Wolfram Language has the means to analyze a poem for such things as meter, rhyme, and cadence. The project above is proof that this can be done. I have recently started rewriting the project to improve it and remove its reliance on custom neural networks (because I don't have the skill to manage them).
If we could drop a stanza into a computer and have it show us the rhythms, echoes, and pauses, then we could see a large part of the craft that went into making the verse (some more than others). Often the little bits of instruction we give to students about rhyme schemes and meter leave them with the impression that writing poetry is like making waffles because the poet has to somehow pour words into a preexisting mould. Having a visualization would allow students to see that the beat and the echoes come from the words themselves, and that the poet is creating the patterns, not following them.
Thanks, Gizem, for making me aware of the Journal of Humanistic Mathematics. In a few months I should be finishing up the current iteration of the project. I'll consider submitting an article at that time.
