Providing online operational support for distributed, security sensitive electronic business processes

Abstract

Online process mining techniques are increasingly used to provide operational support. In this work we describe tools to support distributed business processes which handle sensitive data and require a high level of security together with real-time validation. The techniques presented here have been specifically developed for real-time compliance checking of distributed processes in choreographies of heterogeneous entities. Challenges include the fast aggregation, analysis and validation of process logs that are collected from the distributed participants. The autonomy of the participating entities has to be respected and no sensitive data pertaining to the content of the individual transactions must be accessed for process support and validation purposes. A validation authority for process monitoring and validation is set up. Together with software agents dispatched to the participating entities the validation authority collects events in a central log and then analyzes these events using a particular representation of the process in form of a validation tree to detect and resolve anomalies. We describe the application of these technologies in a distributed business process with more than 400,000 daily process executions. The business process is supported by a help desk managing and responding to incidents and anomalies. We observe a reduction of 90% of the calls to the help desk and an average reduction of 15% in call length. Further the help desk was enabled to act pro-actively, calling participants to the process even before they became aware of anomalies that affected their organization

Laying the ground for future cross-organizational process mining research and application: a literature review

“…Also, the patients’ waiting times could be reduced, which led to higher patient satisfaction. Talamo et al. (2015), in contrast, evaluated their approach to a distributed process involving 400,000 daily executions and a service desk or support.…”

“…To minimize execution time, the remaining tasks of running process instances are allocated to different servers. Based on the concept of validation trees (see Section 4.2.5.3), Talamo et al (2015) explore running service tickets and present a solution, where a validation authority automatically resolves issues, recommends actions or notifies a service agent. Furthermore, three articles apply prediction techniques: Ho et al (2009) use artificial neural networks to predict the quality of currently produced workpieces based on a real-time analysis of quality characteristics.…”

Systematic review on process mining and security

“…A novel system for the provision of efficient operational support for distributed and security sensitive business processes in which automated process validation and extensive troubleshooting functionality is closely integrated with IT Service Desk (SD) operations has been presented [41]. That system's capabilities enable the efficient and automated detection and resolution of anomalies in distributed business processes [41]. The system improves Service Desk (SD) performance and increases process security through real time compliance checking [41].…”

“…That system's capabilities enable the efficient and automated detection and resolution of anomalies in distributed business processes [41]. The system improves Service Desk (SD) performance and increases process security through real time compliance checking [41]. The system further reduces security risks associated with SD operations and SD staff-user interactions including social engineering attacks as it allows SD staff to cross-reference user provided information with system reports [41].…”

“…The system improves Service Desk (SD) performance and increases process security through real time compliance checking [41]. The system further reduces security risks associated with SD operations and SD staff-user interactions including social engineering attacks as it allows SD staff to cross-reference user provided information with system reports [41]. 4) In the Fraud detection category the identified papers describe the new detection methods of Process-based Fraud (PBF) and implementations.…”