Improving Patient Outcomes by Listening to Their Social Media Communications

It is difficult to understand the effectiveness of various treatment options when a huge number of external factors such as lifestyle, diet, and environment affect the burden of a disease. A major barrier to understanding is the challenge of scale—sampling enough patients to separate the major, minor, and negligible factors. With access to a database of more than one trillion public social media posts this project will start addressing this problem by studying the public online communication of patients with skin cancer, specifically basal cell carcinoma, using tools from natural language processing, machine learning, and sentiment analysis.