Tell Me What to Do: Automatically Generating Process Improvement Suggestions

Date

2025-01-07

Contributor

Advisor

Department

Instructor

Depositor

Speaker

Researcher

Consultant

Interviewer

Narrator

Transcriber

Annotator

Journal Title

Journal ISSN

Volume Title

Publisher

Volume

Number/Issue

Starting Page

5638

Ending Page

Alternative Title

Abstract

Process mining techniques play an important role for understanding, analyzing, and improving business processes. Despite their value, deriving actionable improvement measures from process mining insights remains challenging, requiring manual analysis by process analysts. Existing approaches and frameworks provide abstract suggestions, necessitating translation into actionable solutions. Recent efforts focus on generating alternative execution paths rather than explainable improvement suggestions based on specific identified weaknesses, leaving process improvement a labor-intensive task. Addressing this gap, we propose a natural language-driven technique leveraging Large Language Models (LLMs) and social media posts as a rich information source for business-to-consumer (B2C) processes. Our technique identifies process weaknesses from social media posts and generates improvement suggestions using multiple knowledge resources. An evaluation against manually annotated posts demonstrates the effectiveness of our approach, producing suggestions perceived as more useful than human-generated ones. Each suggestion is traceable to its source, enhancing explainability and validity. Furthermore, our technique allows to adapt its knowledge base, allowing seamless integration of additional knowledge resources. Thus, it offers a promising avenue to automate and streamline process redesign efforts across diverse contexts, reducing manual effort in the business process management lifecycle.

Description

Keywords

Business Process Technology, automatic improvement, business process improvement, large language models (llms)

Citation

Extent

10

Format

Geographic Location

Time Period

Related To

Proceedings of the 58th Hawaii International Conference on System Sciences

Related To (URI)

Table of Contents

Rights

Attribution-NonCommercial-NoDerivatives 4.0 International

Rights Holder

Local Contexts

Email libraryada-l@lists.hawaii.edu if you need this content in ADA-compliant format.