Page 358 -
P. 358
340 14 Epilogue
clearly define concepts and possible attribute values. Logging formats such as XES
and SA-MXML (cf. Chap. 4) can relate event data to ontologies. However, the chal-
lenge is to make sure that organizations actually start using semantically annotated
event logs.
Another challenge is produce process models that have a quality and understand-
ability comparable to geographic maps. As shown in Chap. 13, we can learn many
lessons from cartography.
Process mining can be used off-line and online. For off-line process mining, only
historic (“post mortem”) data is needed and no tight coupling between the process
mining software and existing enterprise information systems is needed. For online
process mining (e.g., providing predictions and recommendations), operational sup-
port capabilities need to be embedded in enterprise information systems. From a
technological point of view this may be challenging. It is difficult to embed such ad-
vanced functionality in legacy systems. Moreover, online process mining typically
requires additional computing power. It is important to overcome these challenges
as the value of operational support based on process mining is evident (cf. Chap. 9).
For example, a process model showing the current status of running cases is much
more interesting than a static process model not showing any “live data”.
14.3 Start Today!
As demonstrated in this book, process mining can be brought into play for many
different purposes. Process mining can be used to diagnose the actual processes.
This is valuable because in many organizations most stakeholders lack a correct,
objective, and accurate view on important operational processes. Process mining
can subsequently be used to improve such processes. Conformance checking can be
used for auditing and compliance. By replaying the event log on a process model,
it is possible to quantify and visualize deviations. Similar techniques can be used to
detect bottlenecks and build predictive models. Given the applicability of process
mining, we hope that this book encourages the reader to start using process mining
today.
The threshold to start an off-line process mining project is really low. Most or-
ganizations have event data hidden in their systems. Once the data is located, con-
version is typically easy. For instance, software tools such as ProMimport, Nitro,
XESame, and OpenXES support the conversion of different sources to MXML or
XES. The freely available open-source process mining tool ProM can be down-
loaded from www.processmining.org. ProM can be applied to any MXML or XES
file and supports all of the process mining techniques mentioned in the preceding
chapters. After reading this book, installing the software, and extracting the event
data, the reader is able experience the “magic” of process mining, i.e., discovering
and improving processes based on facts rather than fiction.