Process mining is a technology that helps to analyse process data, so-called event logs, in order to
- discover the process that is actually used (Discovery)
- check whether the used process conforms to the modeled or defined process (Conformance)
- improve the process based on information about its actual performance, e.g. removing bottle necks (Enhancement).
For these tasks, a number of new procedures and algorithms are used, sometimes developed specifically for process mining, and sometimes taken from related areas such as data mining or machine learning.
An overview of the usage of process mining of or as part of software processes is given in my paper (SO20, in German language). You can download this paper here.
If you are interested in using process mining in your company, in particular in the context of software processes, the please contact me. I will be glad to support you.
Most tools for process mining, for example ProM, use event logs in the eXtensible Event Stream (XES) format as their starting point. In order to convert event logs from other formats such as CSV or Excel (XLS), I have written a small Excel VBA application XLS2XES that can be downloaded here. (Version as of 2015-05-23. If you already use the previous version then please update to this new, much improved version.)
If you have any feedback, bug reports or improvement suggestions, please contact me under firstname.lastname@example.org.