Business Process Management 2023 in Utrecht – Highlights and Numbers
From 11th till 15th of September, the 21st International Conference on Business Process Management unfolded in Utrecht, The Netherlands. The BPM conference is the annual key event where academic researchers and industry experts gather to explore the world of business process management. Each year, this event is a fertile ground for displaying innovative research and best practices. Through engaging presentations, tutorials and discussions, attendees gain a deeper understanding of key topics and share experiences. The conference brings together scientists and practitioners from a variety of disciplines including computer science, information systems engineering, and management.
SAP Signavio had a presence at the conference, where we show-cased several demo applications in the fields of business process simulation and declarative conformance checking. These demo applications were the result of cooperation with the Technical University of Munich and the University of Mannheim. In addition, we are actively engaged in pioneering research such as NLP and process automation effects. Also, SAP Signavio co-authored the paper on conversational process modelling in collaboration with the Technical University of Munich.
Almost a dozen of SAP Signavio team members participated at the Utrecht conference, and afterwards they shared their favourite works:
- Agent Miner: An Algorithm for Discovering Agent Systems from Event Data by Andrei Tour, Artem Polyvyanyy, Anna Kalenkova, Arik Senderovich. This paper provides first evidence that considering the resource dimension (i.e., the agent executing the actual work) can lead to better process discovery results. We are excited about this work because we know that integrating business processes – artificial abstractions that allow managing goal-driven operations in organizations – with data representing the perspective of the people and systems that actually execute the work is crucial – that is why we launched our journey-to-process analytics initiative!
- Large Language Models can accomplish Business Process Management Tasks by Michael Grohs, Luka Abb, Nourhan Elsayed, Jana-Rebecca Rehse. The study presented uses a large language model, to perform three BPM tasks: mining imperative process models, mining declarative process models, and assessing RPA suitability from textual descriptions. This approach leverages LLM’s ability to perform these tasks with performance as good as or better than specialized task-specific applications This is interesting because it demonstrates the quality of the LLM’s work and raises the question of whether most of the BPM studies published in the last decade are outdated or whether they really have a raison d’être.
- Towards a Theory on Process Automation Effects by Hoang Vu, Jennifer Haase, Henrik Leopold, Jan Mendling. The paper focuses on how humans perceive automation technology when working within a process, suggesting an effective engagement model between technology, process participants, process managers, and software developers. It covers the human dimension in RPA initiatives, emphasizing that alongside automation efficiency, it is crucial to maintain a great user experience for customers and users.
- Process improvement topic is resonating and promising. The colleagues found these works very insightful:
- From Automatic Workaround Detection to Process Improvement: A Case Study by Nesi Outmazgin, Wouter van der Waal, Iris Beerepoot, Irit Hadar, Inge van de Weerd & Pnina Soffer. The paper proposes a streamlined end-to-end approach that attempts to leverage workarounds to improve processes.
- Not Here, But There: Human Resource Allocation Patterns by Kanika Goel, Tobias Fehrer, Maximilian Röglinger, Moe T. Wynn. The paper presents different human resource allocation patterns for performance improvement from a resource perspective.
BPM in Numbers
The conference witnessed the presentation of 76 papers, contributed by a diverse pool of 201 authors. These authors hailed from 19 countries, representing 83 universities and companies. The papers covered an extensive range of topics, categorized into five distinct areas, as depicted in the graph below.
Within the realm of process management, the scientific community is compact and centres around a handful of key players. The graph below displays these influential figures and the number of papers they contributed to the conference.
As mentioned above, nearly 100 institutes that contributed to research efforts. With an average publication count one per institute, several business process management centres have emerged as prominent contributors. The absolute leaders by number of publications are RWTH Aachen University, Eindhoven University of Technolgy, and TU Munich. Eindhoven University of Technology holds a special place in the history of business process management, serving as one of its origins.
|Organization||Number of papers|
|RWTH Aachen University||12|
|Technical University of Munich||8|
|Eindhoven University of Technology||8|
|Free University of Bozen-Bolzano||5|
Collaboration reaches a significant level in research activities. The graph visually represents the percentage of papers involving international collaboration and partnerships among universities. Specifically, 34% of these papers demonstrate international cooperation, while a substantial 53% signify collaboration between universities. There are 17% of papers which were created together with different private organizations such as SAP Signavio, Celonis, Bonitasoft, etc. This indicates close internal ties within the community.
A schematic depiction of university connections is presented here, illustrating collaborative partnerships in the context of publishing scientific papers for conferences. In this visualization, pink nodes represent universities with substantial affiliations, having established five or more connections with other organizations in the context of conference paper collaborations. Blue nodes represent those with moderate affiliations, typically ranging from two to five connections, while green nodes signify universities with only one such connection. This results in the emergence of distinct “clusters” where ten universities demonstrate strong collaboration with others. The newness of business process management can explain this trend as a field of research. Initially, BPM gained momentum as influential researchers of that time transitioned between different universities. As the field continued to develop, it naturally found its roots independently in various other academic institutions.
Below is a more detailed European map displayed.
Keywords of articles give us a good understanding of the general trend in the area. In total, there are 236 unique keywords. The word cloud visually represents the frequency of keywords, with “process mining” being the most prominent.
By manual analysis of the abstracts, huge interest is noticeable in artificial intelligence (AI), natural language processing (NLP), and machine learning. Though not reflected in keywords, these topics are emerging, indicating alignment with current trends in computer science.
Attendees at the conference expressed their delight in the opportunity to interact with us from a research standpoint, sharing their own findings with our experts and learning from our research and practical experiences. Such kind of organic, genuine engagement holds significant value.
All in all, the BPM Conference stands as a vibrant platform for exchanging knowledge and ideas for the business process management community. While the community may be relatively small, it exhibits steady growth. Additionally, the prospect of monitoring publishing trends for the conference over time and their alignment with broader developments in computer science is an interesting prospect for the future analysis.