Vacancy for two PhD researchers in Predictive and Prescriptive Process Modelling

You will work at the LIRIS research group of the Faculty of Economics and Business at KU Leuven. One position is hosted at KU Leuven, with an exchange period at University of Melbourne. The second position is hosted from the University of Melbourne and includes an exchange period at KU Leuven.
Website unit

Project

Project 1 : Prescriptive Business Process Modelling (hosted at University of Melbourne)

This project will design and evaluate methods for automatically constructing process models that dictate the optimal execution of future business processes, aiming to enhance their efficiency and effectiveness. It will design and evaluate innovative algorithms that ensure that the recommended process actions consistently result in improved overall process outcomes. Unlike current prescriptive process monitoring approaches, which typically offer guidance at the individual process case level, such as intervening at a timely moment to steer a single process case toward success, the constructed prescriptive process models will comprehensively outline the necessary steps for ensuring favorable outcomes across all future business process executions. To achieve this goal, this project will adopt causal analysis that follows Pearl's Causal Hierarchy framework, including the identification of causal relationships between process activities, resources, and data, the application of the derived causal knowledge to plan effective process interventions, and the explanation of the recommended interventions by conducting what-if analysis and retrospective reasoning. The results of this project will significantly enhance business operations, leading to increased competitiveness, profitability, and sustainability of organizations.

Project 2: Predictive Business Process Modelling (hosted at KU Leuven)

This research project pioneers advanced algorithms for predictive business process modelling, specifically focusing on forecasting multidimensional process models. Unlike traditional methods that only predict control flow sequences, our approach extends to forecasting interactions with resources, data objects, decision logic, and performance metrics like bottlenecks. At the core of our strategy is the development of specialized neural networks designed to learn from a novel representation called the process knowledge graph. This representation offers a comprehensive view of process dynamics beyond sequential activities. Our goal is to introduce this innovative representation alongside a Graph Neural Network-based prediction algorithm, enabling businesses to forecast process knowledge graphs and gain proactive insights into their operations. Additionally, we aim to develop simulation models capable of utilizing forecasted process knowledge graphs for optimization. These models will empower decision-makers with what-if analyses, allowing them to optimize future process states. By integrating predictive capabilities with simula tion-driven optimization, our project aims to bridge the gap between foresight and action in business process management. In summary, this research contributes to advancing business process modeling by providing practical tools for decision-making in real-world settings. Through innovative algorithmic development and the integration of predictive and simulation-based optimization techniques, we aim to empower businesses to navigate complex operational landscapes confidently and efficiently.

Profile

Candidates preferably have a master's degree in Business and Information Systems Engineering, Business Engineering, Informatics, Computer Science, Machine Learning, Artificial Intelligence, Information Management, Statistics or related discipline. Excellent (honors-level or better) results in prior studies are required. Candidates must satisfy the prerequisites for admission to the PhD programme of our faculty. There is a strict requirement that you can demonstrate academic excellence (at least honours level) for at least two years. For international candidates in particular, a GRE or GMAT result above the 75th percentile on the quantitative part and an English TOEFL (minimum score 575 paper-based, 233 computer-based, 90 internet-based), or IELTS (minimum score 7) test, both not older than 5 years are required to enter the program. In addition, we require:

• a strong mathematical background
• solid Python programming skills
• good background knowledge in the areas of data science, machine learning, statistics, and process modelling
• passion for research, willingness to go the extra mile, creativity
• ability to work efficiently in a research setting, i.e. be able to investigate new research questions and solutions.

Offer

We offer a 1-year renewable bursary contract, for up to 4 years. 

Interested?

For more information please contact Prof. dr. Jochen De Weerdt, tel.: +32 16 37 62 68, mail: jochen.deweerdt@kuleuven.be or Prof. dr. Johannes De Smedt, tel.: +32 16 37 20 45, mail: johannes.desmedt@kuleuven.be.

KU Leuven strives for an inclusive, respectful and socially safe environment. We embrace diversity among individuals and groups as an asset. Open dialogue and differences in perspective are essential for an ambitious research and educational environment. In our commitment to equal opportunity, we recognize the consequences of historical inequalities. We do not accept any form of discrimination based on, but not limited to, gender identity and expression, sexual orientation, age, ethnic or national background, skin colour, religious and philosophical diversity, neurodivergence, employment disability, health, or socioeconomic status. For questions about accessibility or support offered, we are happy to assist you at this email address.