This project documents how I web-scrape all U.S. congressional speeches on the 2019 Hong Kong protest movements and applied BERTopic models to classify the individual texts into different topics.
-
Webscrape: Web scraping with Selenium, on Congress.gov
-
BERTopic: Apply Bidirectional Encoder Representations from Transformers (BERT) models to classify the topics of each speech.