PROJECT SUMMARY

Evaluating Automated Cultural Curating and Ranking Systems with Synthetic Data

Focus Areas: News & Media
Status: Active

As culture becomes increasingly mediated through automated systems, it is crucial to devise means of evaluating their performance from a societal perspective. Automated systems increasingly navigate a sea of content on our behalf, shaping the flow of culture and cultural products such as news, information retrieval and search, recommender systems and music and video.

This project addresses the question of how to study and evaluate the role that automated systems play in shaping this flow. It brings together technical innovations in the use of synthetic data for hybrid approaches to simulate a broad range of behaviours that influence automated decision-making systems, with a specific focus on ranking-based systems.

It will build tools and approaches to evaluate how recommender and ranking-based systems might incorporate cultural and civic values, such as better representation of artists of colour on streaming platforms or higher visibility and circulation for information of significant public interest.

This project will develop strategies for intervening in the automated flow of culture to advance Centre priorities of responsive, ethical, and inclusive automated decision making.

PROJECT OBJECTIVES

  • Understand what cultural products people are exposed to and in what order and why;
  • Evaluate whether and how ranking-based systems might incorporate factors and values that range from commercial to cultural and civic;
  • Understand how different automated systems create an overall cultural milieu through their combinations and interactions with one another;
  • Provide key insights into how ADMs are experienced in specific sites;
  • Develop ways to ensure that automated intervention in cultural flows aligns with social priorities, including commitments to diversity, fairness, and inclusion; and
  • Demonstrate the appropriate contexts and methods for using synthetic data in evaluating ADM systems.

MORE INFORMATION

RESEARCHERS

Mark Andrejevic

Prof Mark Andrejevic

Project Co-Leader and Chief Investigator,
Monash University

Learn more

Jeffrey Chan

Assoc Prof Jeffrey Chan

Project Co-Leader and Associate Investigator,
RMIT University

Learn more

Kylie Pappalardo profile picture

Dr Kylie Pappalardo

Project Co-Leader and Associate Investigator,
QUT

Learn more

James Meese

Assoc Prof James Meese

Project Co-Leader and Chief Investigator,
RMIT University

Learn more

ADM+S Associate Director Jean Burgess

Prof Jean Burgess

Chief Investigator,
QUT

Learn more

Prof Flora Salim

Prof Flora Salim

Chief Investigator,
UNSW

Learn more

ADM+S Chief Investigator Mark Sanderson

Prof Mark Sanderson

Chief Investigator,
RMIT University
Learn more

Distinguished Professor Julian Thomas

Prof Julian Thomas

Chief Investigator,
RMIT University
Learn more

Danula Hettiachchi

Dr Danula Hettiachchi

Associate Investigator,
RMIT University
Learn more

Joel Stern

Dr Joel Stern

Associate Investigator,
RMIT University
Learn more

Patrik Wikstrom

Prof Patrik Wikstrom

Associate Investigator,
QUT

Learn more

Kyle Herbertson

Kyle Herbertson

Affiliate,
Monash University

Learn more

Xinye Wanyan

Xinye Wanyan

PhD Student,
RMIT University

Learn more

PARTNERS

ABC logo

Australian
Broadcasting
Corporation

Visit website

NYU Ultra Violet Logo

Centre for Responsible AI
New York University

Visit website