Verifying Resource Compliance Requirements from Natural Language Text over Event Logs

Henryk Mustroph, Marisol Barrientos, Karolin Winter, Stefanie Rinderle-Ma

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

Process compliance aims to ensure that processes adhere to requirements imposed by natural language texts such as regulatory documents. Existing approaches assume that requirements are available in a formalized manner using, e.g., linear temporal logic, leaving the question open of how to automatically extract and formalize them for verification. Especially with the constantly growing amount of regulatory documents and their frequent updates, it can be preferable to provide an approach that enables the verification of processes with requirements in natural language text instead of formalized requirements. To this end, this paper presents an approach that copes with the verification of resource compliance requirements, e.g., which resource shall perform which activity, in natural language over event logs. The approach relies on a comprehensive literature analysis to identify resource compliance patterns. It then contrasts these patterns with resource patterns reflecting the process perspective, while considering the natural language perspective. We combine the state-of-the-art GPT-4 technology for pre-processing the natural language text with a customized compliance verification component to identify and verify resource compliance requirements. Thereby, the approach distinguishes different resource patterns including multiple organizational perspectives. The approach is evaluated based on a set of well-established process descriptions and synthesized event logs generated by a process execution engine as well as the BPIC 2020 dataset.

Original languageEnglish
Title of host publicationBusiness Process Management - 21st International Conference, BPM 2023, Proceedings
EditorsChiara Di Francescomarino, Andrea Burattin, Christian Janiesch, Shazia Sadiq
PublisherSpringer Science and Business Media Deutschland GmbH
Pages249-265
Number of pages17
ISBN (Print)9783031416194
DOIs
StatePublished - 2023
EventProceedings of the 21st International Conference on Business Process Management , BPM 2023 - Utrecht, Netherlands
Duration: 11 Sep 202315 Sep 2023

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume14159 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

ConferenceProceedings of the 21st International Conference on Business Process Management , BPM 2023
Country/TerritoryNetherlands
CityUtrecht
Period11/09/2315/09/23

Keywords

  • Compliance Requirements Verification
  • Event Logs
  • Natural Language Text
  • Process Descriptions
  • Resource Mining

Fingerprint

Dive into the research topics of 'Verifying Resource Compliance Requirements from Natural Language Text over Event Logs'. Together they form a unique fingerprint.

Cite this