We used the Process Mining PM2 method (Van Eck, Lu, Leemans, & Van Der Aalst, 2015), which is a<br>simpler and more flexible adaptation of other PM methods such as the L*Life-cycle model (Van Der Aalst, 2011).<br>The PM2 method is structured into four stages (Figure 2): (1) extraction - the data is extracted from the Information bases (Coursera in our case), (2) event log generation– the table value information is modeled in terms<br>of event logs, defining the concepts of case (execution of a process), activities (steps of the process), and temporal<br>order of the activities, (3) model discovery– process mining discovery algorithms are applied to the event log in<br>order to automatically mine a process model describing the observed behaviour of the process, and (4) model<br>analysis– the discovered process models are analyzed in order to understand the observed behaviour. This method<br>was selected because it is the one used in disciplines such as healthcare and business to understand users’ interactive<br>workflows within a particular system (Arias-Chaves & Rojas-Cordoba, 2014; Rojas, Munoz-Gama, Sepúlveda, &<br>Capurro, 2016). It is also suitable for the analysis of both structured and unstructured processes (Van Eck et al.,<br>1<br>Coursera
đang được dịch, vui lòng đợi..
