README

What is this repository for?

• Sentiment/Bigram Analysis and Topic Modeling of the SOTU speeches using NLP tools

Dependencies

This code runs on Python version 3 and above. It also uses libraries like nltk, sklearn, pylmnn, seaborn

How to run the code###

• Run sentiment.py - Assign each speech a "sentiment" score and identify important events from the results
• Run heatmap.py - Find similarity between speeches using KL Divergence
• Run topic.py - Identify the most common topic of a speaker and the top 3 topics for each speech

Detailed Analysis

For more details check out: https://sites.google.com/view/ananyamukherjeehome/data-science/sotu-analysis