TweetyPy
- This application allows a client to use a predictive model to determine which user is more likely to have tweeted a given text
- Developed framework using Flask Python that queries the Twitter API for tweets from various users
- Implemented word2vect using a SpaCy NLP model to create embeddings from the tweet text
- Stored embedded tweet data in a SQLAlchemy Database
- Fit Scikit-Learn Logistic Regression model to tweet data to make predictions, serializing the results for online use


twitter.py Natural Language Processing

