Exploiting a large data base by Longman

  • Authors:
  • A. Michiels;J. Mullenders;J. Noël

  • Affiliations:
  • University of Liège, Belgium;University of Liège, Belgium;University of Liège, Belgium

  • Venue:
  • COLING '80 Proceedings of the 8th conference on Computational linguistics
  • Year:
  • 1980

Quantified Score

Hi-index 0.00

Visualization

Abstract

We wish to explore some of the aspects of the exploitation of two dictionary files by LONGMAN Ltd, one for 'core' English and one for English idioms.We'll try to show the feasibility of an approach to language processing based on a lexicon, conceived of as the repository of grammatical, semantic and knowledge-of-the-world information.After giving a brief description of the computer files (Section I) we'll focus on the following points:(a)a lexical approach to grammar allows a considerable simplification of the PSG component of a parsing system (Section II, Part One);(b)the syntactic potential of many lexemes (at surface structure level) can serve as a guide to their deep structure configurations (Section II, Part Two);(c)provided that a dictionary makes use of a limited defining vocabulary, the texts of the dictionary definitions can be processed on the basis of correlations between syntactic structures (filled with individual lexemes or lexemes belonging to specifiable classes) and semantic relationships such as that between a process verb and an instrument (Section III).