Evaluates ChatGPT's recommendation capabilities beyond accuracy, focusing on fairness, diversity, novelty, and personalization. Includes code, datasets, and evaluation scripts for analyzing LLM-based recommender systems.

Accepted at ACM RecSys 2025

Repository containing the code of SofAgent, the Multi-Agent System realized in collaboration with Natuzzi Italia Innovation Center that leverages a multi-agent approach to enhance the Natuzzi customers' experience. Submitted to IIR 2025

Universal Behavioral Modeling Data Challenge

Official GitHub repository of the lecture "Enhancing Reproducibility in Recommender Systems: A Path Towards Scientific Integrity and Effective Implementation", at the 2024 ACM RecSys Summer School

Framework for detecting label noise in longitudinal Alzheimer’s datasets using Explainable AI (XAI). Combines SHAP-based temporal analysis with subject-level classification to identify and flag potentially mislabeled diagnostic progressions.