Correlation mining: Mining process orchestrations without case identifiers

Abstract

Process discovery algorithms aim to capture process orchestration models from event logs. These algorithms have been designed for logs in which events that belong to the same case are related to each other - and to that case - by means of a unique case identifier. However, in service oriented systems these case identifiers are usually not stored beyond request-response pairs, which makes it hard to relate events that belong to the same case. This is known as the correlation challenge. This paper addresses the correlation challenge by introducing a new process discovery algorithm, called the correlation miner, that facilitates process discovery when events are not associated with a case identifier. Experiments performed on both synthetic and real-world event logs show the applicability of the correlation miner.