Project-4

Git Repository for Metis Project -4

Inspiration behind this project:

Analyze the history of America by analyzing the trends from the speeches of Presidents. I performed topic modeling and then walked back in time to connect the topics with the relevant historical context.

Key questions:

• Which all topics I could model based on the data.
• Polarity/Subjectivity analysis of the President Speeches.

Files :

• Project-4.ipynb - This has all the code that I have used as part of project -4.
• Metis Project -4_Final.pdf - Presentation.

Tools :

• Pandas
• Tableau
• IBM ToneAnalyzer V3

Data Source :

The data has been sourced from Kaggle :https://www.kaggle.com/littleotter/united-states-presidential-speeches. It has the presidential speeches until September 9th, 2019. There are around 992 speeches which range from 1789-04-30 to 2019-09-25 .

Key Analysis Points:

• Topics Modeling
• Sentiment Analysis

Reference:

• https://datayo.org/p/stories/8t0a_WslB_NkaYwz/unpacking-russian-presidential-speech-patterns-with-machine-learning
• https://www.expert.ai/blog/president-biden-inaugural-address-a-linguistic-analysis
• https://www.mosaicdatascience.com/2016/05/09/nlp-development-analyzing-debate-speeches/
• https://towardsdatascience.com/us-presidential-voices-over-the-ages-35678518f82b