Mobile and Social Computing Lab @ FBK

Computational Social Science


We develop computational approaches to the understanding of individual and collective human behaviors using the newly ubiquitous sources of data that are becoming available about our daily life (e.g., mobile phone data, credit card transactions, social media, etc.). In particular, we develop approaches for modeling human mobility, predicting urban crime and unemployment rates, quantifying segregation in cities, etc. 

Geometric Deep Learning and Graph Neural Networks

We develop innovative theoretical results and algorithmic approaches on the fields of equivariant neural networks, temporal graph neural networks, etc. Additionally, we are working on evaluating and developing explainable methods for graph neural networks.

Cooperative AI


We develop and evaluate social agents based on large language models, we define innovative cooperative schemas (inspired by collective intelligence and social learning mechanisms) for coordinating multiple AI agents, for example, multiple deep reinforcement learning agents), etc.