Data-Driven Insights on Human Activity and Social Phenomena

Filippo Simini, University of Bristol
The ALCF's Simulation, Data, and Learning Workshop

Abstract: The analysis of large datasets on human activities can give us precious insights into some of the major societal challenges of our times, from urban sprawl to shrinking cities, from recovery after natural disasters to the impact of human travels on our health and the environment. In this talk I will present various examples of applied research driven by data related to human activity and human behaviour. I will show how scaling theory can help defining a principled methodology to classify human settlement patterns extracted from a high-resolution geospatial database on the distribution of built-up areas globally. I will present a statistical methodology to estimate in real-time the recovery status of small businesses in urban areas hit by natural disasters using public data related to social media usage. Finally, I will give an overview of modelling approaches to generate realistic individual trajectories and collective mobility flows starting from location data collected by cell phones and GPS devices.