Auflistung BISE 61(6) - December 2019 nach Autor:in "Depaire, Benoît"
1 - 2 von 2
Treffer pro Seite
Sortieroptionen
- ZeitschriftenartikelGenerating Artificial Data for Empirical Analysis of Control-flow Discovery Algorithms(Business & Information Systems Engineering: Vol. 61, No. 6, 2019) Jouck, Toon; Depaire, BenoîtWithin the process mining domain, research on comparing control-flow (CF) discovery techniques has gained importance. A crucial building block of empirical analysis of CF discovery techniques is obtaining the appropriate evaluation data. Currently, there is no answer to the question of how to collect such evaluation data. The paper introduces a methodology for generating artificial event data (GED) and an implementation called the Process Tree and Log Generator. The GED methodology and its implementation provide users with full control over the characteristics of the generated event data and an integration within the ProM framework. Unlike existing approaches, there is no tradeoff between including long-term dependencies and soundness of the process. The contributions of the paper provide a solution for a necessary step in the empirical analysis of CF discovery algorithms.
- ZeitschriftenartikelTowards Confirmatory Process Discovery: Making Assertions About the Underlying System(Business & Information Systems Engineering: Vol. 61, No. 6, 2019) Janssenswillen, Gert; Depaire, BenoîtThe focus in the field of process mining, and process discovery in particular, has thus far been on exploring and describing event data by the means of models. Since the obtained models are often directly based on a sample of event data, the question whether they also apply to the real process typically remains unanswered. As the underlying process is unknown in real life, there is a need for unbiased estimators to assess the system-quality of a discovered model, and subsequently make assertions about the process. In this paper, an experiment is described and discussed to analyze whether existing fitness, precision and generalization metrics can be used as unbiased estimators of system fitness and system precision. The results show that important biases exist, which makes it currently nearly impossible to objectively measure the ability of a model to represent the system.