Evaluation Beyond Relevance - Accounting for Credible, Correct and Fair Information

Watch a 78-minute lecture exploring how to evaluate information systems beyond basic relevance metrics, focusing on credibility, correctness, and fairness. Learn from Maria Maistro, a tenure track assistant professor at the University of Copenhagen, as she examines quantitative evaluation methods for detecting threats like misinformation and bias in search and recommendation systems. Discover approaches for developing evaluation measures that can identify misinformation, analyze various fairness metrics and their limitations, and understand the implications of recent EU AI regulations. Gain insights into the challenges of machine learning-based personalization systems and explore promising future directions for creating more reliable and equitable information retrieval solutions.