BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Learning\, Representing\, and Understanding Language - Aida Nematz
 adeh\, DeepMind
DTSTART:20181019T110000Z
DTEND:20181019T120000Z
UID:TALK108109@talks.cam.ac.uk
CONTACT:Andrew Caines
DESCRIPTION:Language is one of the greatest puzzles of both human and arti
 ficial intelligence (AI). Human children learn and understand their langua
 ge effortlessly\; yet\, we do not fully understand how they do so. Moreove
 r\, although access to more data and computation has resulted in recent ad
 vances in AI systems\, they are still far from human performance in many l
 anguage tasks. In my research\, I try to address two broad questions: how 
 do humans learn\, represent\, and understand language? And how can this in
 form AI?\n\nIn the first part of my talk\, I show how computational modeli
 ng can help us understand the mechanisms underlying child word learning. I
  introduce an unsupervised model that learns word meanings using general c
 ognitive mechanisms\; the model processes data that approximates child inp
 ut and assumes no built-in linguistic knowledge. Next\, I explain how cogn
 itive science of language can help us examine current AI models and develo
 p improved ones. In particular\, I focus on how investigating human semant
 ic processing helps us model semantic representations more accurately. Fin
 ally\, I explain how we can use experiments in theory-of-mind to examine q
 uestion-answering models with respect to reasoning capacity about beliefs.
LOCATION:FW26\, Computer Laboratory
END:VEVENT
END:VCALENDAR