Journal of Information and Knowledge (Formerly SRELS Journal of Information Management)

Open Access Open Access  Restricted Access Subscription Access

Error Analysis in ChatGPT’s MARC21 Records: A Study of RDA Conformity


Affiliations
1 Department of Studies in Library and Information Science, University of Mysore, Mysuru – 570005, Karnataka, India

   Subscribe/Renew Journal


This study aims to evaluate the accuracy and quality of MARC21 catalogue records generated by ChatGPT using bibli-ographic data from title pages. The study will shed light on the effectiveness and reliability of automated cataloguing processes utilising AI technology. This involves examining factors such as correctness, consistency, and adherence to Resource Description and Access (RDA) standards. The analysis highlights variations in error rates across different records. Identifying the underlying causes of errors in records with higher rates can help in implementing targeted improvements to enhance the data quality and consistency of ChatGPT-generated catalogue records.

Keywords

Artificial Intelligence, Cataloguing, ChatGPT, MARC21, Resource Description and Access, Quality of Cataloguing
User
About The Authors

B. Niveditha
Department of Studies in Library and Information Science, University of Mysore, Mysuru – 570005, Karnataka
India

N. S. Harinarayana
Department of Studies in Library and Information Science, University of Mysore, Mysuru – 570005, Karnataka
India

Chikku Balachandran
Department of Studies in Library and Information Science, University of Mysore, Mysuru – 570005, Karnataka
India


Notifications

Abstract Views: 139




  • Error Analysis in ChatGPT’s MARC21 Records: A Study of RDA Conformity

Abstract Views: 139  | 

Authors

B. Niveditha
Department of Studies in Library and Information Science, University of Mysore, Mysuru – 570005, Karnataka, India
N. S. Harinarayana
Department of Studies in Library and Information Science, University of Mysore, Mysuru – 570005, Karnataka, India
Chikku Balachandran
Department of Studies in Library and Information Science, University of Mysore, Mysuru – 570005, Karnataka, India

Abstract


This study aims to evaluate the accuracy and quality of MARC21 catalogue records generated by ChatGPT using bibli-ographic data from title pages. The study will shed light on the effectiveness and reliability of automated cataloguing processes utilising AI technology. This involves examining factors such as correctness, consistency, and adherence to Resource Description and Access (RDA) standards. The analysis highlights variations in error rates across different records. Identifying the underlying causes of errors in records with higher rates can help in implementing targeted improvements to enhance the data quality and consistency of ChatGPT-generated catalogue records.

Keywords


Artificial Intelligence, Cataloguing, ChatGPT, MARC21, Resource Description and Access, Quality of Cataloguing



DOI: https://doi.org/10.17821/srels%2F2024%2Fv61i4%2F171481