FAI

Fairness in Artificial Intelligence — June 27, 2022 @ Bocconi University

Location: Bocconi University, Velodromo, Room N04
(Velodromo is a circular red-brick building at the center of the Bocconi main campus. Entering the campus from Via Bligny it is straight ahead. Entering the campus from Via Roentgen 1, turn left at the Gud coffee shop)

Talks will adress: the legal framework around bias and discrimination; bias in natural language processing; bias in image processing; and bias in predictions. The panel will feature the perspective of regulatory bodies, companies, and society.

The workshop is open to the public. Attendance is free but we require pre-registration:

• Registration form

Organizer: Luca Trevisan (Bocconi)
Local arrangements: Diana Righi (Bocconi)

[Zoom Link for remote attendance]

Program:

• 9:30-9:55. Welcome reception
• 9:55-10:00. Opening remarks

• 10:00-10:45. Oreste Pollicino (Bocconi): Algorithmic bias and legal perspectives. The constellation of rights taken seriously [slides]
• 10:45-11:30. Symeon Papadopoulos (ITI): AI Bias in Visual Data [slides]
• 11:30-11:45. Coffee break
• 11:45-12:30. Panel discussion
  • Panelists: Brando Benifei (EU Member of Parliament), Alessandro Bonaita (Generali), Luisella Giani (Oracle), Cinzia Maiolini (CGIL), Massimo Airoldi (Università di Milano)
  • Moderator: Damiano Terziotti (EDGE)
• 12:30-14:00. Lunch break
• 14:00-14:45. Sanghamitra Dutta (JP Morgan): Fairness under feature exemptions: Counterfactual and observational measures [slides]

• 14:45-15:30. Debora Nozza (Bocconi): Roadmap to universal hate speech detection [slides]
• 15:30-16:00. Coffee break
• 16:00-16:45. Omer Reingold (Stanford): Algorithmic Fairness and Loss Minimization [slides]
• 16:45-17:30. Dirk Hovy (Bocconi): Five Sources of Biases in Natural Language Processing (and What to Do about Them) [slides]
• 17:30-18:15. Cynthia Dwork (Harvard): The Calculus of Inclusion [slides]

Part of the initiative "A+I: Algoritmi + Inclusivi" by EDGE.

Supported by the ERC Advanced Grant "SO-ReCodi", funded by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreements N. 834861)