For three years I assisted Tahir Hemphill on his Rapalmanac project (2015-2018). This involved the analysis of over 300,000 hip hop songs, using tools such as IBM Watson to assess sentiment and language-level.
The current project can be found at rapalmanac.com
I worked on Tahir Hemphill’s Rap ALmanac Project between 2014-2018, and am currently training my replacement. The project seeks to create a comprehensive collection of all hip hop lyrics, regardless of language; analyze the data; and create an API to access the data and the analyses.
NOTE: The https certificate for the site has expired! I’ll get Brandon (the current site admin) to fix that ASAP!!!!
I imported and analyzed over 300,000 songs using a variety of natural language libraries. Analyses include language level, sentiment, sophistication and rhyme type. Perhaps the most difficult part of the project has been cleaning up the data so that lyrics in dozens of different languages can co-exist harmoniously.
Photo of Tupac graffiti by Cat Branchman from Seattle, U$A – 2Pac, CC BY 2.0, https://commons.wikimedia.org/w/index.php?curid=45941417