Augmenting cataloguers
planning an AI Agent to generate MARC21 records
Keywords:
Metadata enhancement, AI, cataloguing automation, Large Language Models (LLM), Retrieval Augmented Generation (RAG)Abstract
This article outlines the planning and development of an AI-powered agent designed to assist with generating records at Manchester Metropolitan University’s Library and Cultural Services. The team explores the use of Large Language Models (LLMs) and Retrieval Augmented Generation (RAG) to automate and enhance cataloguing processes, particularly for unique and multilingual materials in the Special Collections Museum and Manchester Poetry Library.
The article discusses technical challenges, ethical considerations, and the importance of maintaining human oversight to ensure quality and transparency. It also details the system architecture and outlines future goals such as multilingual support and automated record enhancement. This project not only aims to improve efficiency but also empowers staff through upskilling and deeper engagement with emerging AI technologies.
References
Albada, Michael (2025) Building Applications with Agents. O'Reilly. Available at: https://learning.oreilly.com/library/view/building-applications-with/9781098176495/ [Accessed: 20 May 2025]
Gonuguntla, H., Meghana, K., Prajwal, T.S., Koundinya, N.V.S.S., Sai, K.N. and Jain, C. (2024) ‘Evaluating Modern Information Extraction Techniques for Complex Document Structures’, International Conference on Electronic, Computer and Energy Technologies (ICECET 2024), Sydney, 25-27 July 2024. Available at: https://doi.org/10.1109/ICECET61485.2024.10698618 [Accessed: 30 April 2025]
Google (2024) What are AI hallucinations? Available at: https://cloud.google.com/discover/what-are-ai-hallucinations [Accessed: 28 May 2025]
Huang, Yu (2024) Levels of AI Agents: from Rules to Large Language Models. Available at: https://doi.org/10.48550/arXiv.2405.06643 [Accessed: 28 May 2025]
Hugging Face (2025) Leaderboards and Evaluations. Available at: https://huggingface.co/docs/leaderboards/en/index [Accessed: 28 May 2025]
Kirmer, Stephanie (2024) Choosing and Implementing Hugging Face Models. Available at: https://www.stephaniekirmer.com/writing/choosingandimplementinghuggingfacemodels [Accessed: 28 May 2025]
Langchain (2025) Langgraph. Available at: https://langchain-ai.github.io/langgraph/ [Accessed: 19 May 2025]
Mendelevitch, Ofer and Bao, Forrest (2025) ‘Rag vs. fine-tuning’ in O. Mendelevitch and F. Bao Hands-On RAG for Production. Available at: https://learning.oreilly.com/library/view/hands-on-rag-for/9798341621701 [Accessed: 20 May 2025]
Resnik, Philip (2024) Large Language Models are Biased Because They Are Large Language Models. Available at: https://doi.org/10.48550/arXiv.2406.13138 [Accessed: 23 May 2025]
Taulli, Tom (2024) What is LangGraph? O'Reilly. Available at: https://learning.oreilly.com/videos/what-is-langgraph/0642572071776/ [Accessed: 20 May 2025]
Tay, Aaron (2025) ‘Deep Dive into Three AI Academic Search Tools’, Katina, (20 May). Available at: https://doi.org/10.1146/katina-052025-2 [Accessed: 20 May 2025]
Ubl, Malte (2020) ‘Design Docs at Google’, Industrial Empathy, 6 July. Available at: https://www.industrialempathy.com/posts/design-docs-at-google/ [Accessed: 28 May 2025]
Urban, Richard (2024) ‘Keeping up with next-generation metadata in archives and special collections’, Hanging Together, 17 December. Available at: https://hangingtogether.org/keeping-up-with-next-generation-metadata-in-archives-and-special-collections/ [Accessed: 8 May 2025]
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Sheldon Korpet, Nathalie Rees
This work is licensed under a Creative Commons Attribution 4.0 International License.
