DATS6312 - Analyzing Presidential Rhetoric and Political Polarization

About

This project analyzes U.S. presidential speeches to study how rhetoric, emotion, and political tone have evolved over time. Using Natural Language Processing (NLP) methods, the goal is to measure sentiment, identify key topics, and explore whether language has become more unifying or divisive across different presidents and political parties.

---

Methods

• Sentiment Analysis: Transformer,
• Topic Modeling: Latent Dirichlet Allocation (LDA), Gensim, NMF
• Classification: TFIDF Logistic Regression, LSTM , DistilBERT
• Interpretability: SHAP, LIME (Future Implementation)

---

Dataset

• Primary Source: American Presidency Project

---

Tools

Python, pandas, numpy, nltk, spacy, scikit-learn, gensim, tensorflow, matplotlib, seaborn, pyLDAvis, shap, lime

---

Authors

Yonathan Shimelis
M.S. Data Science
The George Washington University

Sayan Patra
M.S. Data Science
The George Washington University

---

License

This project is licensed under the MIT License — see the LICENSE file for details.

---