Towards a benchmark dataset for large language models in the context of process automation

The field of process automation possesses a substantial corpus of textual documentation that can be leveraged with Large Language Models (LLMs) and Natural Language Understanding (NLU) systems. Recent advancements in diverse LLMs, available in open source, present an opportunity to utilize them effe...

Full description

Saved in:
Bibliographic Details
Main Authors: Tejennour Tizaoui, Ruomu Tan
Format: Article
Language:English
Published: Elsevier 2024-12-01
Series:Digital Chemical Engineering
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2772508124000486
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850248187341701120
author Tejennour Tizaoui
Ruomu Tan
author_facet Tejennour Tizaoui
Ruomu Tan
author_sort Tejennour Tizaoui
collection DOAJ
description The field of process automation possesses a substantial corpus of textual documentation that can be leveraged with Large Language Models (LLMs) and Natural Language Understanding (NLU) systems. Recent advancements in diverse LLMs, available in open source, present an opportunity to utilize them effectively in this area. However, LLMs are pre-trained on general textual data and lack knowledge in more specialized and niche areas such as process automation. Furthermore, the lack of datasets specifically tailored to process automation makes it difficult to assess the effectiveness of LLMs in this domain accurately. This paper aims to lay the foundation for creating a multitask benchmark for evaluating and adapting LLMs in process automation. In the paper, we introduce a novel workflow for semi-automated data generation, specifically tailored to creating extractive Question Answering (QA) datasets. The proposed methodology in this paper involves extracting passages from academic papers focusing on process automation, generating corresponding questions, and subsequently annotating and evaluating the dataset. The dataset initially created also undergoes data augmentation and is evaluated using metrics for semantic similarity. This study then benchmarked six LLMs on the newly created extractive QA dataset for process automation.
format Article
id doaj-art-e182dad619b44c9ebe9c4949d86c8ebf
institution OA Journals
issn 2772-5081
language English
publishDate 2024-12-01
publisher Elsevier
record_format Article
series Digital Chemical Engineering
spelling doaj-art-e182dad619b44c9ebe9c4949d86c8ebf2025-08-20T01:58:46ZengElsevierDigital Chemical Engineering2772-50812024-12-011310018610.1016/j.dche.2024.100186Towards a benchmark dataset for large language models in the context of process automationTejennour Tizaoui0Ruomu Tan1Tejennour Tizaoui, Technical University of Munich, Chair for Data Processing, Munich, Germany; Corresponding author.Ruomu Tan, ABB Corporate Research Center, Ladenburg, GermanyThe field of process automation possesses a substantial corpus of textual documentation that can be leveraged with Large Language Models (LLMs) and Natural Language Understanding (NLU) systems. Recent advancements in diverse LLMs, available in open source, present an opportunity to utilize them effectively in this area. However, LLMs are pre-trained on general textual data and lack knowledge in more specialized and niche areas such as process automation. Furthermore, the lack of datasets specifically tailored to process automation makes it difficult to assess the effectiveness of LLMs in this domain accurately. This paper aims to lay the foundation for creating a multitask benchmark for evaluating and adapting LLMs in process automation. In the paper, we introduce a novel workflow for semi-automated data generation, specifically tailored to creating extractive Question Answering (QA) datasets. The proposed methodology in this paper involves extracting passages from academic papers focusing on process automation, generating corresponding questions, and subsequently annotating and evaluating the dataset. The dataset initially created also undergoes data augmentation and is evaluated using metrics for semantic similarity. This study then benchmarked six LLMs on the newly created extractive QA dataset for process automation.http://www.sciencedirect.com/science/article/pii/S2772508124000486Large language models (LLMs)Natural language understanding (NLU) process automationExtractive question answering (QA)Natural language processing (NLP)
spellingShingle Tejennour Tizaoui
Ruomu Tan
Towards a benchmark dataset for large language models in the context of process automation
Digital Chemical Engineering
Large language models (LLMs)
Natural language understanding (NLU) process automation
Extractive question answering (QA)
Natural language processing (NLP)
title Towards a benchmark dataset for large language models in the context of process automation
title_full Towards a benchmark dataset for large language models in the context of process automation
title_fullStr Towards a benchmark dataset for large language models in the context of process automation
title_full_unstemmed Towards a benchmark dataset for large language models in the context of process automation
title_short Towards a benchmark dataset for large language models in the context of process automation
title_sort towards a benchmark dataset for large language models in the context of process automation
topic Large language models (LLMs)
Natural language understanding (NLU) process automation
Extractive question answering (QA)
Natural language processing (NLP)
url http://www.sciencedirect.com/science/article/pii/S2772508124000486
work_keys_str_mv AT tejennourtizaoui towardsabenchmarkdatasetforlargelanguagemodelsinthecontextofprocessautomation
AT ruomutan towardsabenchmarkdatasetforlargelanguagemodelsinthecontextofprocessautomation