Page 285 -

10.2 ProM 267

Fig. 10.4 Screenshot of ProM 6. After loading the event log described in Table 7.1,alist of
applicable plug-ins is shown and the plug-in implementing the α-algorithm is selected

ProM 6 (released in November 2010) is based on XES rather than MXML. XES
is the new process mining standard adopted by the IEEE Task Force on Process
Mining (cf. Sect. 4.3). Although ProM 5.2 was already able to load enormous event
logs, scalability and efficiency were further improved by using OpenXES [48, 49].
ProM 6 can distribute the execution of plug-ins over multiple computers. This can
be used to improve performance (e.g., using grid computing) and to offer ProM as
a service. For instance, at TU/e (Eindhoven University of Technology) we use a
dedicated process mining grid to handle huge data sets and to conduct large-scale
experiments [16]. The user interface has been re-implemented to be able to deal
with many plug-ins, logs, and models at the same time. Plug-ins are now distributed
over so-called packages and can be chained into composite plug-ins. Packages con-
tain related sets of plug-ins. ProM 6 provides a so-called package manager to add,
remove, and update packages. Users should only load packages that are relevant
for the tasks they want to perform. This way it is possible to avoid overloading the
user with irrelevant functionality. Moreover, ProM 6 can be customized for domain
specific or even organization specific applications.
Figures 10.4 and 10.5 show the selection of the Alpha miner plug-in and the re-
sulting process model discovered by ProM 6. Figure 10.6 shows a screenshot of the
social network miner of ProM 6 while analyzing interactions between individuals in
a Dutch municipality.
Not all plug-ins of ProM 5.2 have been reimplemented in ProM 6. Neverthe-
less, most of the process mining techniques described in this book are supported
by new plug-ins developed for ProM 6. Table 10.1 shows some of these plug-ins.
These are all related to process mining. However, it should be noted that ProM
(both version 5.2 and 6) supports process analysis in the broadest sense, e.g., also
the analysis techniques mentioned in Sect. 2.3 are supported by ProM or the tools
that ProM interfaces with (e.g., CPN Tools). For example, the plug-in “Analyze
structural properties of a Petri net” computes transition invariants, place invariants,