As a college senior affected by the coronavirus pandemic, I, like over 75% of the students at the University of Texas at Austin, am attending zoom university this fall. Being away from campus during my penultimate semester has made me feel a little homesick. Fueled by curiosity and a little bit of FOMO, I decided to check up on my fellow longhorns with the help of a few Natural Language Processing techniques.
The code for this project is on my GitHub repo.
Table of Contents
- UT Data
- Bag of Words
- Topic Modeling
- Sentiment Analysis
- _1. _Overall Sentiment
- _2. _Sentiment by Topic
- Comparison with Texas A&M
- _1. _Topics in Common
- _2. _Sentiment Comparison
- _3. _School Reputation
- Conclusion
UT Austin Data
I collected 1,348 text entries from the most recent posts on the UT Austin subreddit using Joseph Lai’s Universal Reddit Scraper. These text entries included post title, body, and comments. I then cleaned the text using a VBA script on excel to remove non-ASCII characters.
#machine-learning #reddit #nlp #covid19 #college
