3rd Workshop on
Natural Language Processing for Business Process Management (NLP4BPM)


@ the International Conference on
Business Process Management (BPM), 
in Krakow, Poland, Sept. 2, 2024

Workshop Schedule

9:00 - 10:30 Paper Session 1:

Opening 

Efficient LLM-Based Conversational Process Modeling (pdf)
Julius Köpke and Aya Safan

Leveraging Generative Vision Models for Extracting Process Models from Documents (pdf)
Marvin Voelter, Raheleh Hadian, Timotheus Kampik, Marius Breitmayer and Manfred Reichert

Using Large Language Models to Generate Process Knowledge from Enterprise Content (pdf)
Sandro Franzoi, Maxime Delwaulle, Julian Dyong, Jan Schaffner, Mara Burger and Jan Vom Brocke

11:00 - 12:30 Paper Session 2:

Enhancement of Low-Level Event Abstraction with Large Language Models (LLMs) (pdf)
Edyta Brzychczy, Krzysztof Kluza and Leszek Szała

ProcessLLM: A Large Language Model Specialized in the Interpretation, Analysis, and Optimization of Business Processes (pdf)
Alina Buss, Wolfgang Kratsch, Sebastian Johannes Schmid and Hongyang Wang

Straight outta Logs: Can LLMs overcome preprocessing in Next Event Prediction? (pdf)
Katharina Brennig, Sascha Kaltenpoth and Oliver Müller

Towards a Benchmark for Causal Business Process Reasoning with LLMs  (pdf)
Fabiana Fournier, Lior Limonad and Inna Skarbovsky

14:00 - 15:30 Interactive Session:

Guidelines for writing and reviewing LLM papers for BPM
Participation is open to all who are interested

Closure

Goals & Topics

The NLP4BPM workshop aims to provide a forum for researchers and practitioners to present, discuss, and evaluate how the use of natural language processing (NLP) and large language models (LLMs) can be used to establish new or improve existing methods, techniques, tools, and process-aware systems that support the different phases of the BPM life-cycle.

In this context, natural language and LLMs can play a variety of roles. Among others, they can be used to describe processes in a comprehensible manner, define the meaning of events and activities, and they can provide support for the conduct of process analyses themselves, e.g., as an  interface for process mining or modeling. In addition, it is possible to exploit the process, domain, and programming knowledge of large language models to generate analysis or improvement suggestions.

In the workshop we aim to receive contributions that covers the whole BPM spectrum (foundations, engineering or management aspects of BPM) by welcoming any contribution that considers textual content or meaning for any BPM task or leverages the implicit knowledge of large language models. This includes, but is not limited to:

Types of Contributions

We are open to receive several types of paper submissions to our workshop:

Interactive Session

This year edition will include a plenary discussion in which all participants will be encouraged to jointly talk about a variety of questions and statements on how to further advance the research area. We plan to use several brainstorming techniques for large groups to foster participation and idea generation. Eventually, we might use the insights from this session to establish a manifesto afterwards.

Submission Instructions

Authors are invited to submit original, unpublished research papers. Papers must be written in English and strictly following the Springer LNBIP style. Indicated page limits apply to the total length of the manuscript, including references.

In addition, where possible, an effort should be made to make the results of the papers reproducible. In case LLMs are used, the prompts used to interact with them should be made available as well as other parameters used in the configuration. When possible, evaluations should include at least one open source LLM so that results can be compared in further research. If that is not possible, the version of the closed LLM and dates  when the experiments where performed should be explicitly stated in the manuscript.

Submitted papers will be evaluated according to their rigor, significance, originality, technical quality and exposition, by at least three distinct members of an international program committee. All submissions must be done via EasyChair: https://easychair.org/my/conference?conf=bpm2024

The workshop papers will be published by Springer as a post-workshop proceedings volume in the series Lecture Notes in Business Information Processing (LNBIP). At least one author of each accepted paper must register and participate in the workshop. Registration is subject to the terms, conditions and procedures of the main BPM’23 conference to be found at its website: https://bpm2024.agh.edu.pl/.

Important Dates

All deadlines are set to 23:59 Anywhere on Earth (AoE), GMT+12.

Workshop Organizers

Program Committee

Previous Editions