Sentiment Analysis on Twitter Data
Regarding Dental Issues associated with Opioid Consumption
Senior Capstone Project
This is my senior capstone project, created by myself and my group partners, for our Data Science Senior Capstone class (in lieu of senior thesis). The purpose of the project was to use sentiment analysis on a dataset of raw tweets from Twitter to identify an association between self-reporting of dental issues and self-reporting of opioid consumption. We utilized a BERT model and corpi of keywords for dental issues and opioid consumption, known as medication-assisted treatment (MAT) keywords. Our separation between training data and testing data was based on the publication date of the tweets and the release of a warning by the FDA related to the development of dental issues due to opioid use disorder (OUD). Our project was developed and executed over the 2022-2023 academic year, with feedback and guidance from Dr. Zidian Xie and Dr. Ajay Anand. Our results were presented in front of the class during Presentation Week and was also highlighted during the Senior Design Day hosted by the Goergen Institute for Data Science at the University of Rochester Hajim School of Engineering.