This repository contains all implementation files and experiments conducted for the Extended Dynamic Bayesian Networks introduced in [1].
All code for detecting anomalies [3] and make Predictions [4, 5] in Business Process Logs can be found in this repository.
All experiments in the papers can be reproduced using the files in the Anomalies and Predictions directory in the project.
- Anomalies: Contains all files used for the experiments found in [1] and [3]
- Concept Drift: Contains the files used for the BPI Challenge 2018 [2]
- Data: Data used for the experiments
- Methods: Implementations of all methods used
- Bohmer: Contains our own implementation of the Likelihood Graphs introduced by Bohmer et al in [6]
- Camargo: Contains the slightly adapted implementation used in [8]
- DiMauro: Implementation used in [9]
- EDBN: Contains our implementation of our Extended Dynamic Bayesian Network model
- Lin: Our implementation of the method described in [10]
- Nolle: Contains the original implementations used by Nolle et al in [7], extra libraries needed
- Tax: Adapted implementation used in [11]
- Pasquadibisceglie: Adapted implementation used in [12]
- Premiere: Adapted implementation used in [14]
- SDL: Implementation of our Single Dense Layer architecture [5]
- Taymouri: Adapted implementation used in [13]
- Predictions: Contains all experiments using predictions [4,5]
- Utils: Some extra implementations regarding datastructures, preprocessing and data generation
- Load data using the Logfile class
data_object = Logfile(...)
- Train the EDBN model
model = EDBN.Train.train(data_object)
- Check for anomalies
EDBN.Anomalies.test(data_object, output_file, model, label_attribute, normal_val)
- Load data
data_object = Data.get_data("Helpdesk")
- Select preprocessing settings
settings = Predictions.setting.STANDARD
- Prepare loaded data with selected settings
data_object.prepare(settings)
- Get prediction method
m = Methods.get_prediction_method("SDL")
- Train the model
model = m.train(data_object)
- Test the model
results = m.test(model, data_object)
- Get the score obtained by the model
accuracy = Predictions.metric.ACCURACY.calculate(results)
- Pauwels, Stephen, and Toon Calders. "An Anomaly Detection Technique for Business Processes based on Extended Dynamic Bayesian Networks." (2019)
- Pauwels, Stephen, and Toon Calders. "Detecting and Explaining Drifts in Yearly Grant Applications." BPI Challenge 2018 (2018)
- Pauwels, Stephen, and Toon Calders. "Detecting Anomalies in Hybrid Business Process Logs." Applied Computing Review, Volume 19 Issue 2 Page 18-30 (2019)
- Pauwels, Stephen, and Toon Calders. "Bayesian Network based Predictions of Business Processes." Accepted for BPM Forum 2020
- Pauwels, Stephen, and Toon Calders. "Incremental Predictive Process Monitoring: The Next Activity Case" Accepted for BPM 2021.
- Böhmer, Kristof, and Stefanie Rinderle-Ma. "Multi-perspective anomaly detection in business process execution events." OTM Confederated International Conferences" On the Move to Meaningful Internet Systems". Springer, Cham, 2016.
- Nolle, Timo, et al. "BINet: Multi-perspective Business Process Anomaly Classification." arXiv preprint arXiv:1902.03155 (2019).
- Camargo, M., Dumas, M., Gonz ́alez-Rojas, O.: Learning accurate lstm models ofbusiness processes. In: International Conference on Business Process Management.pp. 286–302. Springer (2019)
- Di Mauro, N., Appice, A., Basile, T.M.: Activity prediction of business processinstances with inception cnn models. In: International Conference of the ItalianAssociation for Artificial Intelligence. pp. 348–361. Springer (2019)
- Lin, L., Wen, L., Wang, J.: Mm-pred: a deep predictive model for multi-attributeevent sequence. In: Proceedings of the 2019 SIAM International Conference onData Mining. pp. 118–126. SIAM (2019)
- Tax, N., Verenich, I., La Rosa, M., Dumas, M.: Predictive business process mon-itoring with lstm neural networks. In: International Conference on Advanced In-formation Systems Engineering. pp. 477–492. Springer (2017)
- Pasquadibisceglie, V., Appice, A., Castellano, G., & Malerba, D. (2019, June). Using convolutional neural networks for predictive process analytics. In 2019 International Conference on Process Mining (ICPM) (pp. 129-136). IEEE.
- Taymouri, F., La Rosa, M., Erfani, S., Bozorgi, Z. D., & Verenich, I. (2020). Predictive Business Process Monitoring via Generative Adversarial Nets: The Case of Next Event Prediction. arXiv preprint arXiv:2003.11268.
- Pasquadibisceglie V., Appice A., Castellano G., Malerba D. (2020) Predictive Process Mining Meets Computer Vision. In: Fahland D., Ghidini C., Becker J., Dumas M. (eds) Business Process Management Forum. BPM 2020. Lecture Notes in Business Information Processing, vol 392. Springer, Cham.