Great idea. Looks like scraping a Kindle library[0] is achievable. I might do something more with it next weekend.

[0] https://github.com/speric/kindle-highlights