GLAM 3: Biases, taxonomies, and participatory machine learning in music information research

GLAM 3: Biases, taxonomies, and participatory machine learning in music information research

A workshop presented by Research axis 2 (Music information research)

Description

Music and sound collections are not only repositories of recordings and artefacts; they are also information systems. Collections and datasets, controlled vocabularies and taxonomies, algorithms, and search interfaces shape what becomes visible, comparable, and researchable, potentially introducing biases and reinforcing stereotypes. These issues resonate across various fields, including information science, libraries and archives, museums and heritage institutions, music technology and musicology, as well as in artistic and research-creation practices that engage with datasets and collections.

This Research Axis 2 workshop at CIRMMT will explore these themes through a series of presentations followed by a panel discussion. The workshop will feature Prof. Rebecca Fiebrink (UAL), whose work on interactive and human-centred ML has influenced contemporary creative practice. She also participated in the UKRI/AHRC “Transforming Collections: Reimagining Art, Nation and Heritage” project, which applied interactive and participatory machine learning to support new ways of engaging with cultural collections across institutions.

Registration

The workshop is free and open to all; however, fill out the registration form by March 5 for food planning.

https://forms.cloud.microsoft/r/HZtLAzaLK0

Schedule

Click here to download the full schedule and abstracts.

12:30–13:15 Light lunch and informal discussions

13:15–13:20 Welcome

13:20–15:20 Individual talks

20 minutes each, followed by 10 minutes of discussion

  • 13:20 Rosemary Mountain – Acknowledging a preference for flexibility and multiple solutions
  • 13:50 Pouya Mohseni – Mapping musical information in the Radif through corpus analysis
  • 14:20 Farzad Daemi Milani – Expanding the tonal vocabulary of music information systems: 17-TET as a framework for recognizing neutral intervals and microtonal harmony
  • 14:50 David Piazza – Materiological biases: strategies for the creative appropriation of implied taxonomies in text-to-audio models

15:20–15:40 Coffee break

15:40–16:20 Rebecca Fiebrink – Interactive ML in the Transforming Collections Project

16:30–17:00 Round table discussion on biases, taxonomies, and participatory machine learning in music information research