Resumen
Daily activities of companies generate and consume massive amounts of data. Different diagrammatic visualizations can be extracted from this data by using different Process Mining algorithms. ProM Framework provides several discovery Process Mining algorithms, mainly focused on the control-flow perspective. This paper analyses the algorithms whose output is either a Process Tree (PT), or an Efficient Process Tree (EPT). The results of several Process Mining algorithms are analyzed and qualitatively evaluated. Precision, Scaled Precision, and Fitness metrics are used for evaluating the resulted diagrammatic visualizations. Moreover, two variations of F-score are also introduced for determining the global quality of the models. The analysis considers, on one hand, two algorithms whose output is a PT and, on the other hand, five versions of an algorithm whose output is an EPT. The findings of this investigation show slightly better results on EPT compared to PT. However, the choice of the most suitable algorithm depends on the analysis type (process discovery, process improvement, audit, risk identification, etc.).