Similarity between pairs of co-indexed trees for textual entailment recognition

Fabio Massimo Zanzotto, Alessandro Moschitti

Research output: Contribution to conferencePaper

Abstract

In this paper we present a novel similarity between pairs of co-indexed trees to automatically learn textual entailment classifiers. We defined a kernel function based on this similarity along with a more classical intra-pair similarity. Experiments show an improvement of 4.4 absolute percent points over state-of-the-art methods.

Original languageEnglish
Pages33-36
Number of pages4
Publication statusPublished - 1 Jan 2020
Event1st Workshop on Graph-Based Algorithms for Natural Language Processing, Textgraphs 2006 at Human Language Technologies - New York City, United States
Duration: 9 Jun 2006 → …

Conference

Conference1st Workshop on Graph-Based Algorithms for Natural Language Processing, Textgraphs 2006 at Human Language Technologies
CountryUnited States
CityNew York City
Period9/6/06 → …

    Fingerprint

ASJC Scopus subject areas

  • Software

Cite this

Zanzotto, F. M., & Moschitti, A. (2020). Similarity between pairs of co-indexed trees for textual entailment recognition. 33-36. Paper presented at 1st Workshop on Graph-Based Algorithms for Natural Language Processing, Textgraphs 2006 at Human Language Technologies, New York City, United States.