Learning Networks of People and Places from Location and Communication Data

Networks and graphs have become essential for understanding the online world. I will discuss how to use mobile location and communication data to build similar networks in the real world offline. We can track movement trends in real time in cities, learn networks of real places, and learn real social networks by gathering long-term high-frequency location data from millions of mobile devices. For example, we can visualize the network of places in a city by showing the similarity between different locations and their activity level in real time. Another graph is the network of users that shows how similar person X is to person Y by comparing their movement histories and how often they colocated. These networks reveal interesting trends in behavior, and they organize people into tribes that are more detailed than traditional demographic groups. With learning algorithms applied to these human activity networks, we can make predictions for advertising, marketing, and social analysis. More importantly, we can build such networks from data without compromising individual privacy, since they only require statistics on user data rather than the original raw bits.