BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:AI Metrics: Theoretical Foundations\, Design\, and Selection of Ev
 aluation Metrics Based on Ground Truth - Enrique Amigó (National Universi
 ty of Distance Education\, Madrid\, Spain)
DTSTART:20260320T120000Z
DTEND:20260320T130000Z
UID:TALK244768@talks.cam.ac.uk
CONTACT:Suchir Salhan
DESCRIPTION:In this talk (based on a book draft\, see this link) I propose
  a unified formal framework for ground truth based evaluation metrics and 
 task characterization grounded in measurement theory. Building on this fou
 ndation\, I analyze the formal properties of existing metrics and organize
  them into families according to task characteristics. The book covers a w
 ide range of discriminative tasks\, including classification\, ranking\, c
 lustering\, and sequence labelling\, among others\, as well as  text gener
 ation. It also provides practical guidance for selecting appropriate metri
 cs depending on the evaluation scenario\, together with a unified software
  framework that implements metrics across multiple tasks. Finally\, the bo
 ok extends evaluation beyond effectiveness to additional dimensions of AI 
 quality\, such as harmfulness\, bias and fairness\, explainability\, and t
 he assessment of cognitive capabilities.\n\n**Bio:** Enrique Amigó is an 
 Assistant Professor at the National University of Distance Education (UNED
 \, Spain) and a member of UNED’s Natural Language Processing and Informa
 tion Retrieval group. His main research interests include evaluation metri
 cs\, document similarity\, representation\, and the connections between In
 formation Access\, Measurement\, Information Theory\, and cognitive scienc
 e. He has received more than 3\,000 citations\, in most cases as first aut
 hor. He has also participated in numerous research projects at the regiona
 l\, national\, and international levels.
LOCATION:SS03 Hybrid (In-Person + Online). Here is the Google Meet Link: h
 ttps://meet.google.com/cru-hcuo-rhu
END:VEVENT
END:VCALENDAR
