eBangor

Towards the emergence of verb-general constructions and early representations for verb entries: Insights from a computational model

Gaspers, J. and Foltz, A. and Cimiano, P. (2014) Towards the emergence of verb-general constructions and early representations for verb entries: Insights from a computational model. In: Annual Conference of the Cognitive Science Society, Quebec, Canada July 23-26, 2014.

Full-text not available from this repository..

Abstract

Recent findings suggest that i) children can build initial verb entries on the basis of syntactic information alone without any additional information provided by a visual context, and ii) that the early representation of verbs encompasses statistical information on the co-occurrence of these verbs with their potential meanings/referents, enabling children to infer verb meanings under referential uncertainty. In this paper we present a computational model that acquires verb-general constructions under referential uncertainty. The model stores linguistic knowledge in line with construction grammar in the form of an interrelated network of constructions. Learning proceeds in line with usage-based theories in an item-based fashion. Computational results show that the model can account for the above-mentioned findings: The model produced patterns similar to those observed in these studies. Our findings hence shed light on the potential mechanisms involved in the emergence of early verb entries and verb-general constructions as well as the representation and refinement of verb entries

Item Type: Conference or Workshop Item (UNSPECIFIED)
Subjects: Research Publications
Departments: College of Arts and Humanities > School of Linguistics and English Language
Date Deposited: 09 Dec 2014 16:28
Last Modified: 23 Sep 2015 02:56
URI: http://e.bangor.ac.uk/id/eprint/217
Publisher: Cognitive Science Society
Administer Item Administer Item

eBangor is powered by EPrints 3 which is developed by the School of Electronics and Computer Science at the University of Southampton. More information and software credits.