Course Review Sentiment Tagging

An application that performs sentiment analysis on course/professor reviews and provides a (+/-) rating and relevant key tags.

Presentation

In Illinois Media Space: https://mediaspace.illinois.edu/media/t/1_n3sykavy

Application

https://share.streamlit.io/rps2ff23/courseproject/main

How to use

1. Enter review text.
2. Click submit!
3. Fetch results
   • Keywords: relevant keywords found in review text (Trained based on RateMyProfessor and Kaggle's Coursera review datasets)
   • Sentiment: one of (strongly negative, negative, positive, strongly positive)
   • Tags: predicted tags for the review, these tags are what are available to select in RateMyProfessor.

File Structure

• /keyword_extract
• /reports: proposal, progress report and final report
• /sentiment_analysis
• /streamlit: code for providing a user interface for the application and deploying the app
• /web_scraping: code for fetching data from RateMyProfessor
• requirements.txt: packages needed for running the app