As information systems are becoming more and more intertwined with the operational processes they support, multitudes of events are recorded by today’s information systems. The goal of process mining is to use such event data to extract process related information, e.g., to automatically discover a process model by observing events recorded by some system or to check the conformance of a given model by comparing it with reality. In this article, we focus on process discovery, i.e., extracting a process model from an event log. We focus on Petri nets as a representation language, because of the concurrent and unstructured nature of real-life processes. The goal is to introduce several approaches to discover Petri nets from event data (notably the a-algorithm, state-based regions, and language-based regions). Moreover, important requirements for process discovery are discussed. For example, process mining is only meaningful if one can deal with incompleteness (only a fraction of all possible behavior is observed) and noise (one would like to abstract from infrequent random behavior). These requirements reveal significant challenges for future research in this domain.
|Title of host publication||Transactions on Petri Nets and Other Models of Concurrency VII|
|Editors||K. Jensen, W.M.P. Aalst, van der, G. Balbo, M. Koutny, K. Wolf|
|Place of Publication||Berlin|
|Publication status||Published - 2013|
|Name||Lecture Notes in Computer Science|
Aalst, van der, W. M. P., & Dongen, van, B. F. (2013). Discovering Petri nets from event logs. In K. Jensen, W. M. P. Aalst, van der, G. Balbo, M. Koutny, & K. Wolf (Eds.), Transactions on Petri Nets and Other Models of Concurrency VII (pp. 372-422). (Lecture Notes in Computer Science; Vol. 7480). Berlin: Springer. https://doi.org/10.1007/978-3-642-38143-0_10