This is a past event
Abstract:
In this talk I will, firstly, review the state of the art of Referring Expressions Generation, viewed as computational models of human reference production. I will ask how successful existing algorithms are from this theoretical perspective, and discuss some recent improvements, for instance relating to the modelling of variation between speakers. Then I will argue that the hardest challenges in Referring Expressions Generation arise from situations in which reference is something other than the "simple" identification of a referent by means of knowledge that the speaker shares with the hearer; I will give examples of the problems that arise in the (frequently arising) more complex situations and of the computational models that address them. -- The talk offers a sneak preview of my upcoming book “Computational Models of Referring: a Study in Cognitive Science”.
- Speaker
- Prof. Kees van Deemter
- Venue
- Meston 311