Privacy-Preserving Directly-Follows Graphs: Supplementary Material

This supplementary material is part of the empirical evaluation performed for the paper entitled "Privacy-Preserving Directly-Follows Graphs: Balancing Risk and Utility in Process Mining." In our empirical evaluation, we study the relation between model parameters representing the disclosure risk, the differential privacy parameter, epsilon, and the utility of the disclosed directly-follows graph (DFG). We conclude that our proposed model optimizes the epsilon value for the input risk and utility metrics. We conducted the experiments using 13 real-life event logs. The source code of our project is part of a prototype called Amun, which is available on GitHub at the following link: https://github.com/Elkoumy/amun The supplementary material contains documentation of the attached files and their column description, and it contains the experimental results as CSV files.This is a supplementary material of the paper entitled "Privacy-Preserving Directly-Follows Graphs: Balancing Risk and Utility in Process Mining.