Semantics-first natural language processing

Authors:
Steven L. Lytinen
Affiliations:
Artificial Intelligence Laboratory, The University of Michigan, Ann Arbor, MI
Venue:
AAAI'91 Proceedings of the ninth National conference on Artificial intelligence - Volume 1
Year:
1991

Citing 5
Cited 0

Information-based syntax and semantics: Vol. 1: fundamentals

Information-based syntax and semantics: Vol. 1: fundamentals
Semantic interpretation and ambiguity

Artificial Intelligence
Understanding Natural Language

Understanding Natural Language
Language As a Cognitive Process: Syntax

Language As a Cognitive Process: Syntax
A knowledge framework for natural language analysis

IJCAI'87 Proceedings of the 10th international joint conference on Artificial intelligence - Volume 2

Quantified Score

Hi-index	0.00

Visualization

Abstract

There is no consensus on how syntax and semantic/ pragmatics should interact in natural language processing. This paper focuses on one issue concerning interaction: order of processing. Two approaches are compared empirically: an interleaved syntax-first approach, in which semantic interpretations is performed at intermediate points during parsing; and a semantics-first approach, in which semantic considerations drive the rule selection process during parsing. The study provides empirical evidence that the semantics-first approach is more efficient than the syntax-first approach in processing texts in narrow domains.