Page 292 -

274 10 Tool Support

of this process mining product less clear. The OKT Process Mining suite by Exeura
uses the process discovery approach presented in [46]. QPR ProcessAnalyzer uses a
process discovery algorithm inspired by the α-algorithm and heuristic mining. The
product has been applied in several Finish hospitals.
Genet, Petrify, Rbminer, and Dbminer are all synthesis tools using state-based
regions [23]. As shown in Sect. 6.4, an event log can be converted into a transi-
tion system and subsequently synthesized into a Petri net. Classical region theory
needs to be extended/relaxed to make it more applicable for process discovery, e.g.,
Rbminer adapts the classical theory to provide more compact and readable process
models [86]. These tools only support control-flow discovery and rely on ProM for
conformance checking [70]. ServiceMosaic is tailored toward the analysis of service
interaction logs, e.g., the logs of HP SOA Manager. The tool discovers transition
systems, but is unable to discover concurrency; instead the focus is on dealing with
noise and protocol refinement [69].
It is impossible to give a complete overview of all products supporting process
mining. Table 10.2 only list products that offer process discovery capabilities. For
instance, we did not list IBM’s WebSphere suite which includes a component called
the WebSphere Business Monitor. IBM aims to support the whole life-cycle using
WebSphere, e.g., when designing a process to be implemented, it is possible to spec-
ify what should be monitored. At run-time KPIs are mapped on the process model
used to configure WebSphere. Note that WebSphere does not support process dis-
covery. As a result there is also no support for the more advanced process mining
techniques. Most of the larger BPM suites, e.g., Global 360, FileNet BPM, Metas-
torm BPM, Oracle BPM Suite, Pegasystems, Savvion BusinessManager, WebMeth-
ods, and Tibco iProcess Suite, still lack mature process mining support. However,
most vendors are currently integrating process discovery capabilities [42] thus fol-
lowing the strategy of Pallas Athena.
Table 10.2 also does not list (a) tools for converting various data sources
into event logs, e.g., ProMimport (promimport.sourceforge.net), Nitro (Fig. 10.9),
and XESame (processmining.org), (b) tools for generating, managing, and com-
paring process models, e.g., PLG [17], Apromore (apromore.org), and BeehiveZ
(sourceforge.net/projects/beehivez/), and (c) process verification/synthesis tools
such as LoLa and Wendy (service-technology.org).

10.4 Outlook

As shown in this chapter, traditional BI products leave much to be desired as they
are not process-centric and focus on simple dashboards and reports rather than more
advanced, i.e. truly “intelligent”, analysis capabilities. Fortunately, there are already
several process mining tools available. ProM supports all of the process mining
techniques discussed in this book. Moreover, there are several commercial tools that
aim at supporting less experienced users or that embed process mining capabilities
in larger systems (cf. BPM|one of Pallas Athena). It seems that most of the larger
BPM vendors are currently embedding process discovery in their products. Few