Published November 10, 2025 | Version v1

PALSYN - Evaluation Result and Models

  • 1. ROR icon German Research Centre for Artificial Intelligence

Description

This repository contains comprehensive evaluation results for the Private Autoregressive Log Synthesizer (PALSYN), a novel approach for generating synthetic event logs with differential privacy guarantees.

Contents

  • Descriptive statistics comparing original and synthetic event logs
  • Generated synthetic event logs with varying privacy parameters
  • Process models extracted from both original and synthetic data
  • Process mining utility measurements across multiple dimension

These evaluation materials complement the PALSYN implementation available at https://github.com/martinkuhn94/PALSYN. 

Files

DETAILED_EVALUATION.zip

Files (685.6 MB)

Name Size
md5:7d64cc8b97991be7ea1a70a3c0869568
607.2 MB Preview Download
md5:2a9c456589078dea1cc3e0247816f632
78.4 MB Preview Download

Additional details

Funding

Federal Ministry of Education and Research
KI-AIM 16KISA115K

Software

Repository URL
https://github.com/martinkuhn94/PALSYN
Programming language
Python
Development Status
Active