Repository logo
 

Unsupervised Entailment Detection between Dependency Graph Fragments

Accepted version
Peer-reviewed

Type

Conference Object

Change log

Authors

Rei, M 
Briscoe, T 

Abstract

Entailment detection systems are generally designed to work either on single words, relations or full sentences. We propose a new task – detecting entailment between dependency graph fragments of any type – which relaxes these restrictions and leads to much wider entailment discovery. An unsupervised framework is described that uses intrinsic similarity, multi-level extrinsic similarity and the detection of negation and hedged language to assign a confidence score to entailment relations between two fragments. The final system achieves 84.1% average precision on a data set of entailment examples from the biomedical domain.

Description

Keywords

Journal Title

Conference Name

Workshop on Biomedical Natural Language Processing

Journal ISSN

Volume Title

Publisher