A Dataset for Entity Recognition in the Automotive Warranty and Goodwill Domain

A large number of warranty and goodwill (W&G) feedback in the automotive sector are collected by global workshop networks. The manual analysis of the feedback texts is a time-consuming procedure. Recent pretrained language models (LMs) finetuned for entity recognition downstream tasks can classi...

Full description

Saved in:
Bibliographic Details
Published in2024 7th International Conference on Artificial Intelligence and Big Data (ICAIBD) pp. 213 - 217
Main Authors Weber, Lukas Jonathan, Ramalingam, Krishnan Jothi, Beyer, Matthias, Liu, Chin, Zimmermann, Axel
Format Conference Proceeding
LanguageEnglish
Published IEEE 24.05.2024
Subjects
Automotive Entity Recognition
BERT
Big Data
Conferences
Data models
Manuals
Task analysis
Text recognition
Transformers
W&G dataset
Warranties
Online AccessGet full text

Cover

More Information
Summary:A large number of warranty and goodwill (W&G) feedback in the automotive sector are collected by global workshop networks. The manual analysis of the feedback texts is a time-consuming procedure. Recent pretrained language models (LMs) finetuned for entity recognition downstream tasks can classify crucial technical terms out of unstructured W & G texts. Annotated W & G-specific entity recognition datasets are necessary for the finetuning of pretrained LMs. However, annotated datasets in the automotive W & G entity recognition domain are not available to the scientific community. We provide the scientific community with the first annotated dataset in the automotive W & G entity recognition domain. Furthermore, we conduct finetuning of pretrained LMs and make them available to the scientific community as well. The annotated dataset and the finetuned models are available via the following link: https://github.com/lukaas95IAutomotive.git
ISSN:2769-3554
DOI:10.1109/ICAIBD62003.2024.10604504