Federico Chesani is employed as assistant professor in the Department of Computer Science and Engineering (DISI), at the University of Bologna, Engineering Faculty. He got his master degree in Informatics and Engineering in 2002; later in 2007 he got his Ph.D. title at the University of Bologna. His research activity is about the adoption of declarative approaches, and in particular of logic programming, to the specification, design and verification of interaction protocols within distributed systems, Multi-Agent Systems (MAS), Web Services and Service Oriented Architectures, and expert systems within the health field. In particular, the research activity of Federico Chesani is focused on the study of formal languages for the definition of interaction protocols, aiming to prove properties, and to verify at run-time the conformance of the peers' observed behavior (possibly detecting wrong situations, and culprits).
Managers and Domain Experts face quite often the challenge of taking decisions on complex matters, characterized by a highly complex scenario, and huge quantity of data and/or knowledge and/or analytical models. Decision Support Systems (DSS) are a class of tools that help the users to understand pros and cons of possible alternative decisions, thus augmenting the awareness of the decision maker about the decision process and its impact.
A first part of the course is devoted to tools and solutions for representing the knowledge and the decision process adopted by Domain Experts. Decision criteria and rules can be documented in a formal, still human-readable way and, more important, can be used to feed software (such as Rule Engines) to fully automatize part of, or whole decision processes. Recent standards like, e.g., the Decision Model and Notation (DMN), will be introduced during a hands-on session, where the audience will directly experiment the advantages of documenting and automatizing decision processes by means of rule-based approaches.
A second part of this course is devoted to introduce the audience to DSS that exploit data and machine learning techniques to learn models of complex domains. These models are in turn used to predict and simulate the scenarios, so as to compare alternative decisions and their expected effects. Visualization techniques then help the decision maker to confront different decisions, w.r.t. optimal solutions for a given objective, or multi-objective decision criteria.