Explore projects

Trained models for extracting semantic process information from event data.

Code for fine-tuning BERT for labeling textual values in events with process-related semantic roles

The repository contains the implementation employed in the paper "Natural Language-based Detection of Semantic Execution Anomalies in Event Logs" by Han van der Aa, Adrian Rebmann, and Henrik Leopold, published in Information Systems (2021)

Multi-Order Concept Drift Detection in Business Processes

This repository includes adjustments made to the original repository to allow (1) the evaluation of implemented change point detection techniques (from cdrift repo) on the CDLG dataset (our data) and (2) on the CDRIFT dataset using relative latency parameters (instead of the absolute value of 200 traces as done in the cdrift repo).

This project provides a prototype of the approach that assesses runtime process flexibility using vector autoregressive modeling