Gerald Penn

A Parsing Algorithm to Reduce Copying in Prolog


Arbeitspapiere des SFB 340, Bericht Nr. 137 (1999), 7pp.
Postscript (89kb)1-up; Postscript gzip-komprimiert (37kb) 1-up , 2-up.

Abstract

This paper presents a new parsing algorithm that reduces the amount of copying required in a memoisation-based Prolog parser for context-free grammars with empty categories to two copies per non-empty edge.


Seminar für Sprachwissenschaft
Eberhard-Karls-Universität Tübingen
Wilhelmstraße 113
72074 Tübingen
Germany