We present a tool for the detection of biased behaviours in AI systems. Using a specific probability-based algorithm, we provide the means to compare the action of the user’s algorithm of choice on a specific feature that they deem “sensible” with respect to fixed classes and to a known optimal behaviour.

BRIOxAlkemy: A Bias detecting tool / G. Coraglia, F.A. D'Asaro, F.A. Genco, D. Giannuzzi, D. Posillipo, G. Primiero, C. Quaggio (CEUR WORKSHOP PROCEEDINGS). - In: BEWARE 2023 : Bias, Ethical AI, Explainability and the role of Logic and Logic Programming / [a cura di] G. Boella, F.A. D'Asaro, A. Dyoub, L. Gorrieri, F.A. Lisi, C. Manganini, G. Primiero. - [s.l] : CEUR Workshop Proceedings, 2024 Jan 12. - pp. 44-60 (( convegno 2nd Workshop on Bias, Ethical AI, Explainability and the role of Logic and Logic Programming co-located with the 22nd International Conference of the Italian Association for Artificial Intelligence tenutosi a Roma nel 2023.

BRIOxAlkemy: A Bias detecting tool

G. Coraglia;F.A. D'Asaro;F.A. Genco
;
G. Primiero;
2024

Abstract

We present a tool for the detection of biased behaviours in AI systems. Using a specific probability-based algorithm, we provide the means to compare the action of the user’s algorithm of choice on a specific feature that they deem “sensible” with respect to fixed classes and to a known optimal behaviour.
Bias; Machine Learning; AI
Settore M-FIL/02 - Logica e Filosofia della Scienza
Settore INF/01 - Informatica
   BIAS, RISK, OPACITY in AI: design, verification and development of Trustworthy AI
   BRIO
   MINISTERO DELL'ISTRUZIONE E DEL MERITO
   2020SSKZ7R_001
12-gen-2024
AI*IA
https://ceur-ws.org/Vol-3615/paper4.pdf
Book Part (author)
File in questo prodotto:
File Dimensione Formato  
paper4.pdf

accesso aperto

Tipologia: Publisher's version/PDF
Dimensione 1.97 MB
Formato Adobe PDF
1.97 MB Adobe PDF Visualizza/Apri
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/2434/1024264
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact