黑龙江大学计算机科学技术学院(软件学院)

学术报告会

Computational Models of Referring

Kees van Deemter

Professor, Dept of Computer Science, University of Aberdeen

时  间：2017年4月26日(星期三)下午2:00-3:30

地  点：黑龙江大学计算机学院会议室(C区4号楼208)

摘  要：

This talk will   introduce the idea of a Natural Language Generation (NLG) system and explore   in more detail one component of such a system, which focusses on referring   expressions. This component decides in what manner to refer to a target   referent, for example, whether to say “the red button”, or “the red button at   the end of the corridor on your right”.

I will introduce some   classic algorithms in this area, discussing what these algorithms are able to   do well and what it is that they still struggle to do. Next, I will use   evidence from experiments with human speakers and hearers to argue that the   most difficult problems in this area arise from situations in which reference   is something other than the "simple" identification of a referent   through shared knowledge; I will give examples of these problematic   situations, and of generation algorithms that address them. To conclude, I   will reflect on the way in which NLG algorithms can embody insights into   human language that other approaches to language have difficulty tackling.

This talk includes   themes from my book “Computational Models of Referring: A Study in Cognitive   Science”, MIT Press, June 2016.

~ All are welcome ~

BIO: I am an academic working in   Computational Linguistics, a research area that belongs to both Artificial   Intelligence and Cognitive Science. My main areas of expertise are   Computational Semantics and Natural Language Generation. I've long taken an   interest in logical and philosophical issues arising from this work; more   recently I've collaborated extensively with psycholinguists interested in   algorithmic models of human language production.

My research centers around   computational models of human communication. One of my specific research   interests is the computational generation of referring expressions, as when   we say 'the inventor of the light bulb', or 'the large icon at the top of   your screen'. I am intrigued by situations in which communication is (or   appears to be) flawed, as when we use expressions that are ambiguous or   vague. Vagueness is the focus of my book "Not Exactly: in Praise of   Vagueness" (Oxford University Press 2010), which has been translated   into both simplified and complex Chinese. My book on referring expressions,   entitled Computational Models of Referring: a Study in Cognitive Science” has   recently appeared with MIT Press.