A Time-Aware Method to Process Behavioral Similarity Calculation

Abstract

Process behavioral similarity calculation is widely used in many activities of business process management, such as process mining, process clustering and process retrieval. However, most existing process behavioral similarity measurement methods do not consider variable constraints (e.g. time constraints) in process models. Since the variable constraints of tasks, especially time constraints, are crucial to quantify the differences between the executions of tasks with the same or similar labels in process models, it is significant to calculate exact process behavioral similarity with the consideration of their impacts in process models. In this paper, we formally define the process behavioral similarity with time constraints and propose a time-aware method to efficiently compute the process behavioral similarity based on the ordering relations with time constraints. Experiments illustrate that our proposal can significantly improve the rationality and accuracy of process behavioral similarity measurement.