Τμήμα: Ιδρυματικά - Δ.Π.Θ.

Ημερομηνία Έγκρισης: 01/04/2024

Επιστημονικά Υπεύθυνος: ΑΘΑΝΑΣΙΑ ΣΥΚΙΩΤΟΥ

Ημερομηνία Λήξης: 31/5/2026

Πλαίσιο Χρηματοδότησης: ΔΙΕΘΝΕΙΣ ΠΟΡΟΙ

Email: asikioto@law.duth.gr

Περιγραφή: The FAIR-PReSONS project, funded by the European Union and coordinated by the University of the Aegean, develops a fair and unbiased AI system for recidivism prediction, with a strong emphasis on gender equality and transparency in criminal justice. By collecting and digitizing data from prisons in Greece, Portugal, and Bulgaria, structuring it into knowledge graphs, and applying advanced algorithms integrated with neural networks, the project enhances both the accuracy and explainability of predictions. Within this initiative, the Democritus University of Thrace (DUTH) plays a key role, drawing on its strong academic expertise in European Union law, criminology, and the legal framework of AI in justice systems. DUTH contributes to data collection, gender and bias analysis, training of legal professionals, and the evaluation of the developed AI system. Through its involvement in FAIR-PReSONS, DUTH strengthens its role in advancing fair and ethical practices in criminal justice, contributing its expertise to ensure unbiased decision-making and gender-sensitive approaches within European justice systems.