Challenges in Service Mining: Record, Check, Discover Wil M.P. van der Aalst 1,2 1 Architecture of Information Systems, Eindhoven University of Technology, P.O. Box 513, NL-5600 MB, Eindhoven, The Netherlands. w.m.p.v.d.aalst@tue.nl 2 International Laboratory of Process-Aware Information Systems, National Research University Higher School of Economics (HSE), 33 Kirpichnaya Str., Moscow, Russia. Abstract. Process mining aims to discover, monitor and improve real processes by extracting knowledge from event logs abundantly available in today’s information systems. Although process mining has been ap- plied in hundreds of organizations and process mining techniques have been embedded in a variety of commercial tools, to date these techniques have rarely been used for analyzing web services. One of the obvious reasons is that cross-organizational event data cannot be shared easily. However, (1) messages exchanged between services tend to be structured, (2) service-orientation continues to be the predominant implementation paradigm, and (3) the most substantial efficiency gains can often only be achieved across different organizations. Hence, there are many possible applications for service mining, i.e., applying process mining techniques to services. If messages are recorded, then one can discover a process describing interactions between services. If, in addition, descriptive or normative models are available, one can use process mining to check con- formance and highlight performance problems. This extended abstract aims to provide pointers to ongoing work on service mining and lists some of the main challenges in this emerging field. 1 From Process Mining to Service Mining Process mining is an enabling technology for service mining. Process mining can be used to discover processes from raw event data, check the conformance of ob- served and modeled behavior, enhance models by improving or extending them with knowledge extracted from event logs [2]. The uptake of process mining is not only illustrated by the growing number of papers, but also by commercial analysis tools providing process mining capabilities, cf. Disco (Fluxicon), Percep- tive Process Mining (Perceptive Software, before Futura Reflect and BPMone by Pallas Athena), ARIS Process Performance Manager (Software AG), Pro- cessAnalyzer (QPR), Interstage Process Discovery (Fujitsu), Discovery Analyst (StereoLOGIC), and XMAnalyzer (XMPro). Web services have become one of the main paradigms for architecting and im- plementing business collaborations within and across organizational boundaries