"Covfefe", or understanding online interactions at scale

Online platforms have enabled people to get together and express opinions at a scale unprecedented in history. Online interactions can range from an inconsequential ‘like’ of a friend’s post or picture to nationally important conversations between citizens and their elected representatives. In this talk I will describe a number of machine learning and data analysis-based approaches to answer three high level questions:  

1. What can social network interactions tell us about ourselves?  

2. How can we co-ordinate disparate and independent user actions to learn new information about the content being generated?  

3. Are online conversations bring us together or driving us apart?