Predicting Logic’s Lyrics with Machine Learning

Using simple machine-learning concepts, we can analyze Logic’s style of hip-hop and predict lyrics he would actually write.

Written for and in Dennis Sun’s DATA 301 class at Cal Poly SLO. Absolutely no permission is given to any person to use or submit any of this code for any reason. I had a fantastic time trying to figure out this lab and simply wish to document and display this work for educational, personal, and employment-related reasons.

Check out the @hanzykamin/predicting-logics-lyrics-with-machine-learning-9e42aff63730">Medium post I wrote about this project!

Language:

• Python 3 (Jupyter Notebook)

Libraries:

• requests
• time
• bs4 (documentation)
• pickle
• random

Tasks:

• Break down all code into gists.
• Make Python scripts for each main section of code.
• Create a Jupyter Notebook.
• Publish the Jupyter Notebook as a Medium post.
• Implement a recurrent neural network instead of a Markov chain.
• Figure out how to generate lines that consistently rhyme with each other.

Please don’t hesitate to either reach out to me or make pull requests with revisions to make this implementation better!