Replication Package for Master Thesis: Replicability of Predictive Process Monitoring Techniques for Outcome Prediction

Replication Packages for "Replicability of predictive process monitoring techniques for outcome prediction" Overview This repository contains the replication packages for my master thesis "Replicability of predictive process monitoring techniques for outcome prediction". Each package provides a complete replication environment including Docker configuration, code, datasets, and results for reproducing the experiments from eight selected studies on predictive process monitoring for outcome prediction. System Requirements - Docker (version 20.0 or higher)- Minimum 8 GB RAM, 50 GB disk space- Linux/Unix environment (tested on Ubuntu 22.04) Package Structure Each package (P1-P8) is provided as a tar.gz file containing:- README.md: Detailed instructions for replicating that study- Dockerfile: Environment setup with exact package versions- scripts: Data preprocessing, hyperparameter tuning, and experiment scripts- datasets: Event logs and preprocessed data- results: Output files and analysis scripts- Result generation scripts for tables and figures in the thesis * Original dataset website sources are documented in Table 19 of the thesis in Appendix A.3. Replicated Studies P1: Leontjeva et al. (2015) - Complex Symbolic Sequence EncodingsP2: Di Francescomarino et al. (2016) - Clustering-Based Predictive Process Monitoring P3: Teinemaa et al. (2019) - Outcome-Oriented Predictive Process Monitoring Review and BenchmarkP4: Teinemaa et al. (2019) - Temporal Stability in Predictive Process MonitoringP5: Márquez-Chamorro et al. (2022) - Updating Prediction ModelsP6: Kim et al. (2022) - Encoding Resource ExperienceP7: Tama & Comuzzi (2022) - Heterogeneous Ensemble LearningP8: Peeperkorn et al. (2023) - Positive and Unlabelled Event Logs * Original repository links for each study can be found in the respective package README.md files. Quick Start Guide 1. Download package: tar -xzf P[X]_**.tar.gz2. Build Docker: docker build -t p[x]_replication .3. Run the container: docker run -it p[x]_replication4. Follow package-specific README.md for detailed execution steps General Execution Pattern Most packages follow this workflow:1. Data preprocessing (converting XES to CSV, labeling)2. Hyperparameter tuning3. Experiments with original metrics4. Extended evaluation with additional metrics5. Results generation and analysis Contact For any questions or issues regarding the materials in this repository, please feel free to contact the author: [e-mail]