Abstract
This article presents a computational model of the production of referring expressions under uncertainty over the hearer's knowledge. Although situations where the hearer's knowledge is uncertain have seldom been addressed in the computational literature, they are common in ordinary communication, for example when a writer addresses an unknown audience, or when a speaker addresses a stranger. We propose a computational model composed of three complimentary heuristics based on, respectively, an estimation of the recipient's knowledge, an estimation of the extent to which a property is unexpected, and the question of what is the optimum number of properties in a given situation. The model was tested in an experiment with human readers, in which it was compared against the Incremental Algorithm and human-produced descriptions. The results suggest that the new model outperforms the Incremental Algorithm in terms of the proportion of correctly identified entities and in terms of the perceived quality of the generated descriptions.
Highlights
A large body of research in psycholinguistics investigates the extent to which speakers tailor their utterances to their addressees, a phenomenon known as audience design (Clark and Murphy, 1982; Clark and Wilkes-Gibbs, 1986b; Isaacs and Clark, 1987; Clark and Brennan, 1991)
Computational models will be employed because they are the most explicit and detailed models of reference production that are currently on the market; controlled experiments with human participants will help us ground our computational model in actual human behavior
The analysis suggests that descriptions produced by the new algorithms have higher “quality” than the ones produced by the Incremental Algorithm (IA) and DBpedia
Summary
A large body of research in psycholinguistics investigates the extent to which speakers tailor their utterances to their addressees, a phenomenon known as audience design (Clark and Murphy, 1982; Clark and Wilkes-Gibbs, 1986b; Isaacs and Clark, 1987; Clark and Brennan, 1991). Referring expressions (REs) are a natural focus for research on audience design, because they aim to identify a referent uniquely for an audience; if the RE includes information unknown to the hearer, the hearer may fail to know what or who the speaker talks about. The link between knowledge and reference makes REs a suitable focus for research on Audience Design. The present article follows this well-trodden path, using computational models, and experiments with human participants. Computational models will be employed because they are the most explicit and detailed models of reference production that are currently on the market (see van Deemter, 2016; Section 2 below); controlled experiments with human participants will help us ground our computational model in actual human behavior
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
