The research activity will mainly focus on the definition of descriptive and predictive models for the analysis and mining of real-time flows of big data from social media and textual data collections in order to study and develop methodologies and algorithms to improve understanding the behavior and relationships emerging from social media, as well as using the enormous amount of user-generated contents to understand in real time the emergence of new events of interest, such as cultural events, but also emergency situations, such as natural disasters or epidemics. For this purpose, supervised and unsupervised techniques of knowledge extraction, sentiment analysis, word embedding, personalized recommender systems, Community Detection and Community Question Answering (CQA) on complex networks, new approaches for data mining and textual documents will be studied and used.
In particular, the activities will be mainly applied towards the definition of clinical decision support models for the diagnosis and treatment of diseases to create systems of early disease diagnosis and treatment prediction through algorithms and diagnostic and predictive techniques of therapeutic response to diseases, also with the support of cyber physical, disease prediction, epidemic outbreaks detection and tracking systems; study of virus and information diffusion phenomena, of resilience and evolution on different types of complex networks such as heterogeneous, multilevel, temporal.
Study and definition of methodologies that combine artificial intelligence and data science techniques for the discovery of patterns and emerging phenomena contained in the huge digitized information published on social media platforms.
Web-based health surveillance, Emergency management, Urban intelligence.