Process analytics through event databases: Potentials for visualizations and process mining

Abstract

Events, routinely broadcasted by news media all over the world, are captured and get recorded to event databases in standardized formats. This wealth of information can be aggregated and get visualized with several ways, to result in alluring illustrations. However, existing aggregation techniques tend to consider that events are fragmentary, or that they are part of a strictly sequential chain. Nevertheless, events’ occurrences may appear with varying structures (i.e., others than sequence), reflecting elements of a larger, implicit process. In this work, we propose several transformation templates to a enable a process perspective for raw event data. The basic idea is to transform event databases into a format suitable for process mining (aka event log) to enable the rich toolbox of process mining tools. We present our approach through the illustrative example of the events that happened in Greece during the referendum period (summer 2015).