From Page to Data: Using AI-Assisted Technologies to Process Archival Collections

Dr. Loren S. Moulds
Head of Digital Scholarship & Preservation and Digital Collections Librarian
University of Virginia School of Law

Loren S. Moulds is the Head of Digital Scholarship & Preservation and Digital Collections Librarian at the University of Virginia School of Law, where he leads the development, interpretation, and management of the Law Library’s digital scholarship projects and oversees the preservation and accessibility of its archival and digital collections. As a member of the University’s General Faculty, Loren collaborates closely with UVA’s digital scholarship and library technology communities to support innovative research and digital initiatives.
 
A historian of twentieth-century American Political Development, Loren received his Ph.D. in History from the University of Virginia in 2014. He is finishing an MSIS with a focus on Digital Archives and Data Management from the University of Tennessee, Knoxville. Loren’s work integrates emerging technologies into historical research and archival practice, enabling new approaches to the transcription, structuring, and analysis of complex historical data. His ongoing research and digital initiatives explore how AI and computational tools can help unlock and interpret archival materials at scale, broadening access to historical collections and fostering new avenues for inquiry.

Description:

As archives continue to digitize vast quantities of historical material, institutions face a persistent challenge: how to process handwritten, unstructured, and inconsistent documents efficiently and at scale. Traditional transcription and data extraction methods are resource-intensive, limiting the accessibility and usability of digital collections. This webinar explores how AI-assisted technologies — particularly large language models (LLMs) like GPT-4 — are transforming this process.

Drawing on real-world applications, this session will walk participants through how LLMs can be used to extract structured data from archival materials that can be handwritten, tabular, and varied. Emphasis will be placed on prompt engineering, iterative refinement, and human-in-the-loop strategies that enable faster, adaptable, and more scalable workflows.

The session will also address how these technologies are reshaping archival pedagogy. It will highlight the importance of equipping LIS and humanities students with the skills to critically engage with AI tools, understand their limitations, and creatively apply them to archival research and data management.

Learning Objectives:

• Understand the potential and limitations of using LLMs for archival transcription and metadata extraction
• Explore practical strategies for applying AI tools to historical collections in various formats
• Learn how to design and refine AI workflows for structured data output and review
• Discuss the pedagogical implications of AI for LIS education and the digital humanities
• Gain insight into integrating these tools ethically and effectively in professional and classroom contexts

Audience:

This webinar is designed for LIS educators and students, digital archivists, metadata librarians, and anyone interested in the intersection of archives, technology, and AI.