"Indigenous and Accountable AI for the Archival Professions"

After this 2-hour interactive workshop, professional archivists and records managers, archival studies students, and archival educators will be able to apply concepts, theory, and methods to critically, ethically, and pragmatically incorporate Artificial Intelligence (AI) and Machine Learning (ML) into their workflows, with a particular focus on upholding Indigenous date sovereignty and governance principles, respecting Indigenous protocols, and using paradata for AI accountability in archival settings. It is also an opportunity to learn about and participate in the development of an evolving curriculum to teach these topics to archival professionals and students.

The workshop will be divided into two sections:

1. Indigenous protocols, data sovereignty and governance, and AI applications

2. Paradata and pragmatics for AI accountability in archives

The goal of this workshop is to explore how archivists can critically evaluate and incorporate new technologies like AI/ML into their work while still upholding their responsibilities as stewards of cultural memory. Furthermore, this workshop also calls attention to the concerns AI integration in archives can present to Indigenous data held in these institutions, aims to provide meaningful and realistic recommendations to addressing these barriers in the archival professions, and proposes pragmatic ways of establishing accountability in uses of AI in archival settings.

Instructors: Alexander Ross, PhD; Kaila Fewster; and Richard Arias-Hernandez, PhD,  University of British Columbia.

Registration is limited to 40 people.

"AI Tools for Audio Archives"

Incomplete metadata is one major barrier to accessibility of audio archives, in this hands-on workshop, we will examine a suite of tools for audio metadata enrichment.

Key tasks include speaker linking (identifying speakers common between audio files), speaker diarization (who is speaking at what times in the recordings), language identification (is the recording in French? English? etc), and speaker indexing and retrieval. Beyond an overview of these tasks and tools, we will also critically examine the strengths and weaknesses of modern approaches to solve these problems.

Participants should bring a laptop, and digital materials will be provided to enable directed practice.

Instructors: Peter Sullivan is a PhD student at the UBC iSchool, whose research interest focus on speech processing, bias and robustness of AI speech models, and applications of AI to audio archives. As part of the InterPARES AI Project, he has been working to improve metadata enrichment of audio archives.  Eng Sengsavang is Reference Archivist at the United Nations Educational, Scientific and Cultural Organization (UNESCO) in Paris, France. She holds dual masters degrees in Archival and Library Studies from the University of British Columbia. Her research as part of InterPARES AI is focused on archives, digitization, and artificial intelligence. She is co-editor with Jens Boel of Recordkeeping in International Organizations: Archives in Transition in Digital, Networked Environments (Routledge, 2021).

Registration is limited to 30 people.