Su, XuanLiu, CongZhang, ShuaipengZeng, QingtianMo, QiCheng, Long2026-04-202026-04-202026-041939-1374PURE: 152718378PURE UUID: 5458d064-9d1a-4e5a-81df-5cc48f3613c2Scopus: 105030678825WOS: 001737537400041http://hdl.handle.net/10362/202395Su, X., Liu, C., Zhang, S., Zeng, Q., Mo, Q., & Cheng, L. (2026). Toward Efficient Support for Business Process Event Log Sampling. IEEE Transactions on Services Computing, 19(2), 1606-1618. https://doi.org/10.1109/TSC.2026.3665370Large volumes of event logs have been accumulated by business information systems. Accompanied by that, various process discovery techniques are invented to uncover underlying business processes based on event logs. Event log sampling, recognized as one of the most effective techniques for accelerating discovery efficiency, has gained significant attention in recent days. However, achieving high performance in sampling while maintaining superior sample log quality remains a challenge for current techniques. To tackle the problem, a novel event log sampling technique, denoted as sigRank, is introduced to improve both the sampling efficiency and the quality of the sample log by quantifying the significance of each trace. The proposed sampling technique has been implemented as a publicly available tool in the open-source process mining platform ProM. Compared with state-of-the-art techniques using 12 public event logs, we experimentally illustrate that the proposed approach can significantly accelerate sampling efficiency while guaranteeing superior sample log quality for process discovery.134405029engEfficiencyEvent Log SamplingPetri NetsProcess DiscoveryQuality EvaluationHardware and ArchitectureComputer Science ApplicationsComputer Networks and CommunicationsInformation Systems and ManagementSDG 9 - Industry, Innovation, and InfrastructureSDG 12 - Responsible Consumption and Productionjournal article10.1109/TSC.2026.3665370https://www.scopus.com/pages/publications/105030678825https://www.webofscience.com/wos/woscc/full-record/WOS:001737537400041