Archives Meet GPT: A Pilot Study on Enhancing Archival Workflows with Large Language Models
Author(s)
Zhang, Shitou
Peng, Siyuan
Wang, Ping
Hou, Jingrui
Issue Date
2024-03-20
Keyword(s)
Archive
Archival work
Large Language Model
ArcGPT
Abstract
Archive management requires meticulous handling and precise stewardship of textual materials. Large Language Models (large language models (LLMs)), trained extensively on text data, possess exceptional text processing and interpretative capabilities. These allow for profound insights and extractions from the vast troves of information within archives. Anchored in the records life-cycle theory and archivists’ practical operations, this research explores the potential advantages of using LLMs in archival work. We begin by constructing a theoretical framework that demonstrates how LLMs can streamline tasks for archivists. Next, we introduce a novel LLM designed specifically for archive work, called Archival Generative Pre-trained Transformer (ArcGPT), and present its initial performance across four archival tasks. Recognizing that overarching performance metrics may not encapsulate the genuine user experience, we further propose a methodology for a user experiment designed to gauge the user-centric performance of how LLMs support archivists in their archival workflows.
Publisher
iSchools
Series/Report Name or Number
iConference 2024 Proceedings
Type of Resource
Other
Language
eng
Handle URL
https://hdl.handle.net/2142/122806
Copyright and License Information
Copyright 2024 is held by Shitou Zhang, Siyuan Peng, Ping Wang, and Jingrui Hou. Copyright permissions, when appropriate, must be obtained directly from the authors.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.
