Inter-organizational business processes involve multiple independent organizations collaborating to achieve mutual interests. Process mining techniques have the potential to allow these organizations to enhance operational efficiency, improve performance, and deepen the understanding of their business based on the recorded process event data. However, inter-organizational process mining faces substantial challenges, including topical secrecy concerns: The involved organizations may not be willing to expose their own data to run mining algorithms jointly with their counterparts or third parties. In this paper, we introduce CONFINE, a novel approach that unlocks process mining on multiple actors' process event data while safeguarding the secrecy and integrity of the original records in an inter-organizational business setting. To ensure that the phases of the presented interaction protocol are secure and that the processed information is hidden from involved and external actors alike, our approach resorts to a decentralized architecture comprised of trusted applications running in Trusted Execution Environments (TEEs). We show the feasibility of our solution by showcasing its application to a healthcare scenario and evaluating our implementation in terms of memory usage and scalability on real-world event logs.

翻译：跨组织业务流程涉及多个独立组织协作以实现共同利益。流程挖掘技术有望使这些组织能够基于记录的事件数据提升运营效率、改善绩效并加深对业务流程的理解。然而，跨组织流程挖掘面临重大挑战，包括主题性机密性问题：相关组织可能不愿将自身数据暴露给合作方或第三方来共同运行挖掘算法。本文提出CONFINE——一种新颖方法，能在跨组织业务场景下，在保护原始记录机密性与完整性的同时，实现对多主体流程事件数据的挖掘。为确保所提出的交互协议各阶段的安全性，并使得参与方及外部各方均无法获取已处理信息，该方法采用由可信执行环境（TEEs）中运行的可信应用程序构成的去中心化架构。我们通过将其应用于医疗场景展示该方案的可行性，并在真实事件日志上从内存使用和可扩展性两方面评估其实施效果。