Assignation of Roles to Individuals/Event Logs: A Graph-Based Approach



Alessandro Berti

Software Engineer


+49 241 80 21912



BSc Thesis Project

Title: Assignation of Roles to Individuals/Event Logs: A Graph-Based Approach

Author: Yosra Ibn Elfekih

Supervisor: Alessandro Berti

1st examiner: Prof.Dr. ir. Wil M.P. van der Aalst

2nd examiner: Prof. Dr.-Ing Ulrik Schroeder


Information systems control most of the business world organizations. The data accumulated is used, managed, and analyzed to provide more information to the organization. Event logs record this kind of data precisely about the execution of business processes and provide a source of information to validate, improve and visualize the actual process executions. The detection and discovery of the roles allows for the enhancement of any business process model. The goal of the thesis is the development of a novel graph-based algorithm of roles discovery. The event logs used are first encoded into event graphs. With the help of the graph embeddings, the features for each node in the event graph are discovered and clustering algorithms will allow afterwards the determination of the similarity between these features. The roles are detected based on the feature similarity. The aim of this thesis is to improve the efficiency and the quality of role discovery, besides the production of a visualization of the discovered roles. The algorithms of this thesis are implemented in the Python Language and the PM4PY library.