报告题目：What next? Modelling dialogical interaction
报告人： Professor Jonathan Ginzburg
报告人单位： Université Paris Diderot (Paris 7)法国巴黎七大
报告人简介： Jonathan Ginzburg, Professor of linguistics at Université Paris Diderot (Paris 7). Professor Ginzburg received Bachelor's Degree in Mathematics in Hebrew University of Jerusalem in 1987, and PhD in Stanford University in Stanford University in 1992. He has worked extensively on the semantics of questions and dialogue. He has written two books ("Interrogative Investigations" and "The Interactive Stance: Meaning for Conversation"), along with over fifty papers in top journals and conference proceedings in semantics, pragmatics, and computational linguistics.
报告摘要： Predicting the next possible move in a dialogue is one of the most challenging problems in cognitive science, closely related to the Turing test. It is also the basis of resolvingfragmentary utterances in dialogue, one of the key differences between text and conversation. In this talk I will discussa theoretical model of this problem formalized within the formalisms of KoS (Ginzburg, 2012) and Type Theory with Records (Cooper, 2012, Cooper and Ginzburg, 2015).Several sub-parts of this model have been tested extensively in corpus work—in particular the range of possible responses to a question with a question (Lupkowski and Ginzburg, 2016) and have been used as the basis for automatic classification of fragmentary utterances in dialogue (Fernandez, Ginzburg, and Lappin, 2007;Dragone and Lison, 2015). I will show how this model can scale up to the macro-structure of conversation and to incremental, word by word processing.
报告题目：Laughter in dialogue
报告人：Ye Tian 博士后
报告人单位：Université Paris Diderot (Paris 7)法国巴黎七大
报告人简介：Ye Tian is a post-doctoral researcher at Laboratoire de linguistique formelle, Université Paris Diderot (Paris 7), in project DUEL: Disfluencies, Laughter and Exclamation in natural dialogues (project DUEL), together with Jonathan Ginzburg, Julian Hough, and David Schlangen. Some of the other topics she has worked on include: Emojis, polar questions and polar particles and contextual relevance (Questions Under Discussion). Prior to her postdoc post, Ye was a PhD student at the linguistics department of UCL (University College London), supervised by Richard Breheny. Her PhD thesis was about negation processing.
报告摘要：A common assumption about laughter is that it is an indicator of the speaker's emotional state (independent of language), we argue that laughter interacts with language and influences propositional content. We propose a semantic/pragmatic account in which laughter is treated as a gestural event anaphor referring to a laughable; the laughable is an event which, upon appraisal, triggers a positive psychological shift in the speaker. This simple meaning of laughter, when combined with rich contexual reasoning, generates a range of functions, such as marking non-literal meaning, softening, acknowledging and scare quoting. Recognizing the function of laughter and producing laughter at appropriate times is a great challenge for a naturalistic dialogue system.