2025
Utilization of LLM for Process Mining Analysis of Event Log of Travel Expenses at the Operational Level
HALAŠKA, Michal and Roman ŠPERKABasic information
Original name
Utilization of LLM for Process Mining Analysis of Event Log of Travel Expenses at the Operational Level
Authors
HALAŠKA, Michal (203 Czech Republic, belonging to the institution) and Roman ŠPERKA (703 Slovakia, guarantor, belonging to the institution)
Edition
1. vyd. London, Utilization of LLM for Process Mining Analysis of Event Log of Travel Expenses at the Operational Level, p. 57-79, 23 pp. AI, Analytics and Strategic Decision-Making, 2025
Publisher
Routledge
Other information
Language
English
Type of outcome
Chapter(s) of a specialized book
Field of Study
10201 Computer sciences, information science, bioinformatics
Country of publisher
United Kingdom of Great Britain and Northern Ireland
Confidentiality degree
is not subject to a state or trade secret
Publication form
printed version "print"
References:
Organization unit
School of Business Administration in Karvina
ISBN
978-1-032-83110-7
Keywords in English
ChatGPT; BPM; LLM; travel expanses; process mining
Tags
International impact, Reviewed
Changed: 2/8/2025 08:47, doc. RNDr. Ing. Roman Šperka, Ph.D.
Abstract
In the original language
This chapter explores the application of large language models (LLMs) in process mining, mainly focusing on their ability to manage complex queries and interpret processes effectively. It aims to identify which business process management (BPM) tasks LLMs can support and the potential transformative impact on BPM workflows. The study uses ChatGPT, a state-of-the-art LLM, to analyze an event log of travel expenses at a university. A series of prompts are designed to evaluate the model’s performance in extracting process descriptions, identifying anomalies, and generating process insights. The findings demonstrate ChatGPT’s proficiency in transforming complex process mining data into intuitive formats, significantly improving the efficiency and precision of process analysis. The model successfully identified process variants, anomalies, and bottlenecks and provided detailed descriptions of process activities. Integrating conversational AI like ChatGPT into process mining can make these tools more accessible and effective, reducing the need for specialized expertise. This integration has the potential to improve traditional process mining techniques, leading to more insightful and actionable results in BPM. This chapter provides a preliminary exploration of the applicability in process mining, offering valuable information on their potential to transform BPM practices. It highlights the innovative use of conversational AI to complement and enhance existing process mining methodologies. The study underscores the need for further research to validate these findings and explore advanced AI technologies for process optimization.