After watching Dan Shiffman's video on using RitaJS to program with text, I was super excited to start creating some sort of visualization with different types of material.
I started off with music lyrics.
I made each button input a different famous lyric from some of the most popular artists on the radio, and a few others. I then used RitaJS to replace certain types of words within the lyrics -- in a sort of mad-lib style -- with other random words either of that same word type or not.
When the code finally worked (which took a long period of playing with the code and seeing what was possible and what wasn't), I finally got to get into the aesthetics and design of the application.
In the future, I would love to create a more in-depth analysis of lyrics. I would love to learn more about the construction of rap lyrics and how they differ within the different popularities of artists. I mean, do artists on the radio typically have the same word structure while underground artists differ? Do artists that begin to have a romantic relationship with another artists, start to create more similar word structure within their music?
Tools that I used:
I really like the design of how you made this feel like an old-school typewriter. It would be interesting to see how the experience might shift if you changed the lexicon you're using for the random word replacement. What if the random words were from other rap songs, or other types of musical genres instead of coming from the more "neutral" RiTa default lexicon? In terms of your other thoughts for analyzing rap lyrics, one direction might be to find common n-grams, as Allison Parish outlines here: http://www.decontextualize.com/teaching/rwet/n-grams-and-markov-chains/, and then markov chains could be another method for poetry generation in the style of rap lyrics.