Societal ai, complex networks dYNAmics, and oPtimization for Sustainable Environments -SYNAPSE

Research Activity

The research activities of the SYNAPSE group are organized into the following areas, integrating the ethical, social, and cultural dimensions of Artificial Intelligence (Societal AI) in a cross-cutting manner:

•        Multimodal Learning & Generative AI: design of architectures and methods for the analysis and understanding of textual, visual, acoustic, and biometric content from heterogeneous sources, including social networks and online platforms. The group develops AI models for synthesis, classification, prediction, and explainability tasks, with particular attention to computational efficiency and environmental impact (Green AI), promoting energy-efficient techniques and continual learning.

•        Health Informatics & Decision Support: development of intelligent systems for clinical diagnosis and decision-making support, including early prediction of neurodegenerative diseases (e.g., Alzheimer’s), treatment response modeling, and epidemic monitoring. Advanced techniques such as graph learning, NLP, and few-shot learning are applied to build predictive and personalized models, ensuring data security and social acceptability of AI solutions.

•        Social Media Analytics: application of topic modeling, sentiment analysis, and event detection techniques to social media data streams. Particular attention is given to identifying communities, emerging dynamics, misinformation,  and multimodal fake news detection, with the goal of promoting responsible AI practices.

•        Complex Networks: development of methods for classification, community detection, and sparsification in complex networks, including structural and functional brain networks. Research explores Continual Learning to address dynamic and incremental scenarios. Advanced applications include network robustness through dual-graph modeling, efficient sparsification, and structural evolution monitoring, with a focus on system resilience, sustainability, and social impact.

•        Mathematical Programming Models: formulation of linear and integer programming models for the management of deterministic or stochastic complex systems with single or multiple objectives.

•        Optimization: Algorithms supported by AI techniques , including exact methods (dynamic programming, branch-and-cut algorithms, cutting planes, Benders decomposition) and heuristic or metaheuristic approaches (local search, metaheuristics, Lagrangian decomposition).

•        Human-Machine Interaction & Trustworthy AI: design of intelligent systems capable of ethical, personalized, and explainable interaction with users. Here, the perspective of Societal AI is central: transparency, inclusiveness, and respect for human values are considered essential in the adoption of AI technologies across social, clinical, and urban contexts.

The group operates within a multidisciplinary synergy, with a strong focus on applied research and experimentation on real-world datasets, maintaining a close connection to the principles of ethical AI, explainability, and the social impact of technology.

Goals

The SYNAPSE research group develops multimodal, sustainable, and human-centered Artificial Intelligence methodologies and architectures to analyze, model, and understand complex and interconnected systems with strong societal impact. It integrates deep learning, natural language processing, generative AI, optimization, graph intelligence, and complex network theory to manage phenomena characterized by heterogeneous data and multi-agent dynamics. Application domains include social networks, digital health, urban mobility, intelligent infrastructures, and cybersecurity. The goal is to design predictive and explainable models for early diagnosis, critical event monitoring, personalized services, and decision support, ensuring transparency, reliability, and fairness in the use of AI for for societal well-being and sustainable innovation.

Application Fields

Cybersecurity and Forensics, Healthcare and Medicine, Emergency Management, Urban Intelligence, Network Optimization, Urban Logistics, Management of Logistics Platforms

Back To Top