Cancer immunotherapy is a promising treatment that harnesses the body's immune system to combat cancer, but not all patients respond to the treatment. To understand patients' views on cancer immunotherapy, we propose a BERTopic Modeling and Sentiment Analysis platform. We were given 4.9 million user-generated posts from various online cancer and health forums and performed text pre-processing on each of them. With 3.6 million cleaned messages, our team chose BERTopic as our methodology to extract the topics people were discussing most frequently. We computed the sentiment of every post in each topic to determine their public perception, and we also compared BERTopic with other methods, showing that it outperformed them. Lastly, we developed a web-based dashboard to support medical professionals and show the topics and their sentiment analysis results through data visualizations. Our proposed platform has the potential to provide valuable insights to medical professionals, giving them a tool to improve patient outcomes and experiences.

• This report represents the work of one or more WPI undergraduate students submitted to the faculty as evidence of completion of a degree requirement. WPI routinely publishes these reports on its website without editorial or peer review.

Creator

• Sheehy, Mago
• Jolibois, Florkenthia
• Wee Sit, Tiffany
• Mangini, Sierra
• Goldinger, Sidney

Publisher

• Worcester Polytechnic Institute

Identifier

• 105436
• E-project-042623-153800

Mot-clé

• Natural Language Processing
• Data Science
• Computer Science
• Cancer Immunotherapy
• Sentiment Analysis
• Big Data

Advisor

• Ngan, Chun-Kit

Year

• 2023

Date created

• 2023-04-26

Resource type

• Major Qualifying Project

Major

• Innovation with User Experience
• Computer Science

Source

• E-project-042623-153800

Rights statement

• In Copyright

Dernière modification

• 2023-06-29