Han van der Aa, Henrik Leopold, Kiran Busch, and Adrian Rebmann
Natural Language Processing (NLP) has become an essential tool for many organizations aiming to analyze and understand the vast amounts of text data they generate. The latest developments related to language models have significantly boosted the analytical capabilities of NLP tools and have created completely new use cases. In this tutorial, we will focus on the intersection of NLP and Business Process Management (BPM) and explore how NLP can support various BPM analysis tasks. We first introduce fundamentals of NLP and explore how language models work. Then, we focus on the automated analysis of textual descriptions, after which we turn to the analysis of process-oriented artifacts, where we show how NLP can be used to obtain novel insights from process models and event logs. These parts are followed by a hands-on exercise session, in which participants will learn how to use general and process-specific NLP libraries and techniques. Finally, we conclude the tutorial with a discussion of future directions. After the tutorial, participants will have learned about the fundamentals of NLP, the potential of using NLP in the context of BPM, and how to apply NLP to their own BPM research and analyses.
Iris Beerepoot, Francesca Zerbato, Barbara Weber, and Pnina Soffer
Despite the rising popularity of process mining in practice, executing a process mining project is a daunting task that requires significant expertise. As such, there is a pressing need to provide comprehensive support for process analysts. This tutorial aims to provide participants with an overview of state-of-the-art practices followed by process analysts at each stage of a typical process mining project, from defining questions to evaluating results. Based on empirical evidence and experience from several projects, we go over concrete strategies to support analysts, with a focus on specific tasks and areas that require extra attention. This sets the stage for further research in developing support for process analysts and allows identifying blind spots that future research might address.
Thomas Grisold, Jan vom Brocke, Wolfgang Kratsch, Jan Mendling, and Maxim Vidgof
Large language models, such as ChatGPT, provide ample opportunities for organizational work. These models are capable of collecting, integrating, and generating information with no or little human supervision. Despite their wide and rapid uptake, we lack systematic knowledge about how large language models can be used in business processes. Our tutorial sheds light on the organizational, managerial and design-related implications of using large language models in business processes. We present a theoretical framework that integrates and synthesizes research from relevant streams, including task complexity, task automation, and human-AI delegation. We specify potential opportunities and threats in relation to various forms of tasks, such as decision tasks and judgment tasks. Along these lines, we also explore how the use of large language models may affect the overall outcome of a process, for example, by providing new value propositions. We use, reflect, and discuss the implications of our framework based on real-world examples. Our conceptual framework is relevant to guide future research but also inform managerial decisions in organizations.