Co-PI: Paolo Giorgini (Università di Trento)
Current AI research has achieved great success by focusing on problems with rather narrow goals (or on complex problems that can be broken down into smaller independent subproblems). To successfully address complexity, however, there is a need for a significant change of pace and to go beyond current research, compartmentalized into separate vertical areas. Specifically, an integrated approach is needed, to which the Spoke’s name refers, Integrative-AI.
The spoke proposes an interdisciplinary approach based on theories and techniques that integrate diversified scientific methods, technologies, disciplines and skills. The spoke will investigate the integration of heterogeneous computational representation and reasoning and learning (inference) techniques, such as the integration of data-driven and model-based approaches, symbolic and sub-symbolic representations.
