Effectively Using Syntax for Recognizing False Entailment
Abstract
Recognizing textual entailment is a challenging problem and a fundamental component of many applications in natural language processing. We present a novel framework for recognizing textual entailment that focuses on the use of syntactic heuristics to recognize false entailment. We give a thorough analysis of our system, which demonstrates state-of-the-art performance on a widely-used test set.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 2006
- Accession Number
- ADA456695
Entities
People
- Arul Menezes
- Lucy Vanderwende
- Rion Snow
Organizations
- Stanford University