A Tabular Method for Island-Driven Context-Free Grammar Parsing

Giorgio Satta, Oliviero Stock

Island-driven parsing is of great relevance for speech recognition/understanding and other natural language processing applications. A bidirectional algorithm is presented that efficiently solves this problem, allowing both any possible determination of the starting words in the input sentence and flexible control. In particular, a mixed bottom-to-top and top-down approach is followed, without leading to redundant partial analyses. The algorithm performance is discussed.

This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.