Operationalizing Responsible AI principles to Music Discoverability AI Tools Development

The subproject focuses on the intersection of cultural policy and responsible AI development. The problem of data bias and
automated recommendation are two known problems in cultural industries and AI development relevant to the Discoverability
project. The project seeks to develop alternative forms of algorithmic recommendation that differ from the de facto emphasis
on personalization. This project answers the need to operationalise principles of Responsible AI at the site of music
discoverability. The intern will produce a Responsible AI Scan (RAI Scan) of the terrain where the classifier and recommender
will be developed and deployed. The RAI scan will provide the context necessary for a precise operationalisation. It will be a
review and synthesis of the literature on responsible AI, expert interviews with stakeholders interpolated or impacted by the
creation of these systems (artists, diversity of cultural content advocates, and other relevant stakeholders). The partner
organization will build and learn from the RAI Scan to conduct the large Discoverability project. It will also be useful for the
future projects of the partner organization

Faculty Supervisor:

Fenwick McKelvey

Student:

Partner:

CEIMIA

Discipline:

Sociology

Sector:

Information and cultural industries; Other services (except public administration); Professional, scientific and technical services

University:

Concordia University

Program: