Scientific talk: Process Mining / Data Science
Target audience: Fits all 3 profiles, but agents/specialists are better suited
Abstract: Process mining is a family of techniques that allows the study of how ideal process relate to actual ones, taking a data-driven point of view, i.e., what actually happens rather than what is supposed to happen. During the session two types of process mining will be discussed: control-flow discovery and conformance checking. Control-flow discovery aims at synthesizing process models from actual executions of them. This allows one to study how processes are actually executed (in-vivo vs in-vitro processes). Conformance checking techniques can help in understanding where real executions deviate from a prescriptive model by identifying process violations as well as locating them.