Happy new year!
Short version: I tried to plot the word frequency used in lyrics. It looks like this:
Just like in literature, the distribution is quite exponential.
You can also find the raw data and the length distribution in the slide show 🙂
How I got the idea:
It was late in the last year, when I was biking through the meadow and listening to the bbc podast about word-frequency and stats analysis in literature, I realised people still interested in research being done years ago, as long as it’s relevant.
And it’s been a while since I read lyrics analysis in various literature in MIR. But the “multi-modal” data analysis was definitely making a come back at ISMIR last year. And in a scala meet up last year, people were analysis lyrics and generating them as well.
It might also become relevant with the dutch folk tunes dataset I’m working on. But how it would come into help could be the material of another post.