Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Process Modeling With Large Language Models

Humam Kourani, Alessandro Berti, Daniel Schuster, Wil M. P. van der Aalst

TL;DR

The paper addresses the difficulty of creating BPM representations from natural language and the barriers non-experts face. It proposes a framework that uses Large Language Models to generate and iteratively refine process representations from textual descriptions, with an intermediate POWL representation and a safe code-generation pipeline, including prompt engineering, error handling, and a feedback loop. The authors implement a concrete system that exports to BPMN and Petri nets while ensuring soundness guarantees through POWL, and demonstrate that GPT-4 achieves strong results compared to Gemini and a textual abstraction baseline. This work demonstrates how generative AI can democratize Business Process Management and streamline modeling tasks, while outlining clear directions for direct BPMN generation and richer interactivity in future work.

Abstract

In the realm of Business Process Management (BPM), process modeling plays a crucial role in translating complex process dynamics into comprehensible visual representations, facilitating the understanding, analysis, improvement, and automation of organizational processes. Traditional process modeling methods often require extensive expertise and can be time-consuming. This paper explores the integration of Large Language Models (LLMs) into process modeling to enhance the accessibility of process modeling, offering a more intuitive entry point for non-experts while augmenting the efficiency of experts. We propose a framework that leverages LLMs for the automated generation and iterative refinement of process models starting from textual descriptions. Our framework involves innovative prompting strategies for effective LLM utilization, along with a secure model generation protocol and an error-handling mechanism. Moreover, we instantiate a concrete system extending our framework. This system provides robust quality guarantees on the models generated and supports exporting them in standard modeling notations, such as the Business Process Modeling Notation (BPMN) and Petri nets. Preliminary results demonstrate the framework's ability to streamline process modeling tasks, underscoring the transformative potential of generative AI in the BPM field.

Process Modeling With Large Language Models

TL;DR

The paper addresses the difficulty of creating BPM representations from natural language and the barriers non-experts face. It proposes a framework that uses Large Language Models to generate and iteratively refine process representations from textual descriptions, with an intermediate POWL representation and a safe code-generation pipeline, including prompt engineering, error handling, and a feedback loop. The authors implement a concrete system that exports to BPMN and Petri nets while ensuring soundness guarantees through POWL, and demonstrate that GPT-4 achieves strong results compared to Gemini and a textual abstraction baseline. This work demonstrates how generative AI can democratize Business Process Management and streamline modeling tasks, while outlining clear directions for direct BPMN generation and richer interactivity in future work.

Abstract

In the realm of Business Process Management (BPM), process modeling plays a crucial role in translating complex process dynamics into comprehensible visual representations, facilitating the understanding, analysis, improvement, and automation of organizational processes. Traditional process modeling methods often require extensive expertise and can be time-consuming. This paper explores the integration of Large Language Models (LLMs) into process modeling to enhance the accessibility of process modeling, offering a more intuitive entry point for non-experts while augmenting the efficiency of experts. We propose a framework that leverages LLMs for the automated generation and iterative refinement of process models starting from textual descriptions. Our framework involves innovative prompting strategies for effective LLM utilization, along with a secure model generation protocol and an error-handling mechanism. Moreover, we instantiate a concrete system extending our framework. This system provides robust quality guarantees on the models generated and supports exporting them in standard modeling notations, such as the Business Process Modeling Notation (BPMN) and Petri nets. Preliminary results demonstrate the framework's ability to streamline process modeling tasks, underscoring the transformative potential of generative AI in the BPM field.
Paper Structure (16 sections, 3 figures, 3 tables)

This paper contains 16 sections, 3 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. LLM-Based Process Modeling Framework
  4. Framework Overview
  5. Process Representation
  6. Prompt Engineering
  7. Model Generation and Refinement
  8. Refinement Loop.
  9. Error Handling
  10. Evaluation
  11. Limitations and Future Directions
  12. Expanding Process Perspectives.
  13. Extended Evaluation and User Studies.
  14. Direct BPMN Generation.
  15. Enhanced Interactivity.
  16. ...and 1 more sections

Figures (3)

  • Figure 1: LLM-based process modeling framework.
  • Figure 2: BPMN models generated for the order handling process in the first run. Although the models generated using our system show some deviations from the original process description, the model generated by GPT-4 correctly captures complex non-hierarchical dependencies. Unlike the models generated using our system, TA led to an unsound model that is dead after the choice between paying and completing an installment agreement.
  • Figure 3: BPMN models generated for the hotel process in the first run. The model generated by GPT-4 using our system provides a high degree of conformance with the process description, significantly surpassing the model generated by Gemini. The model generated using TA is unsound as the end event is not reachable after the second instance of "Readies Cart".