OKCupid Date-A-Scientist: Using Naive Bayes to predict whether a user likes dogs

Hi all! I’m a dog lover, and thought it would be fun to try to predict whether a user likes dogs based on this data. I’m interested in NLP, so I figured I’d do it based on the Essay questions and use Naive Bayes Classification.

Github link: https://github.com/GavinWhelan/OK-Cupid-Date-A-Scientist-Codecademy

I’ve been trying to decide how I might increase the accuracy of this model (~61% using both unigram and bigram models). I would love any feedback, especially on those next steps and on the general presentation/readability of the code!