Distributional models for lexical semantics: An investigation of different representations for natural language learning

Danilo Croce, Simone Filice, Roberto Basili

Research output: Contribution to journalArticle

Abstract

Language learning systems usually generalize linguistic observations into rules and patterns that are statistical models of higher level semantic inferences. When the availability of training data is scarce, lexical information can be limited by data sparseness effects and generalization is thus needed. Distributional models represent lexical semantic information in terms of the basic co-occurrences between words in large-scale text collections. As recent works already address, the definition of proper distributional models as well as methods able to express the meaning of phrases or sentences as operations on lexical representations is a complex problem, and a still largely open issue. In this paper, a perspective centered on Convolution Kernels is discussed and the formulation of a Partial Tree Kernel that integrates syntactic information and lexical generalization is studied. Moreover a large scale investigation of different representation spaces, each capturing a different linguistic relation, is provided.

Original languageEnglish
Pages (from-to)115-134
Number of pages20
JournalStudies in Computational Intelligence
Volume589
DOIs
Publication statusPublished - 2015
Externally publishedYes

Fingerprint

Linguistics
Semantics
Syntactics
Convolution
Learning systems
Availability
Statistical Models

Keywords

  • Distributional lexical semantics
  • Kernel methods
  • Question classification

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Distributional models for lexical semantics : An investigation of different representations for natural language learning. / Croce, Danilo; Filice, Simone; Basili, Roberto.

In: Studies in Computational Intelligence, Vol. 589, 2015, p. 115-134.

Research output: Contribution to journalArticle

@article{b5274b9df3c24f05bc538fbe8881be31,
title = "Distributional models for lexical semantics: An investigation of different representations for natural language learning",
abstract = "Language learning systems usually generalize linguistic observations into rules and patterns that are statistical models of higher level semantic inferences. When the availability of training data is scarce, lexical information can be limited by data sparseness effects and generalization is thus needed. Distributional models represent lexical semantic information in terms of the basic co-occurrences between words in large-scale text collections. As recent works already address, the definition of proper distributional models as well as methods able to express the meaning of phrases or sentences as operations on lexical representations is a complex problem, and a still largely open issue. In this paper, a perspective centered on Convolution Kernels is discussed and the formulation of a Partial Tree Kernel that integrates syntactic information and lexical generalization is studied. Moreover a large scale investigation of different representation spaces, each capturing a different linguistic relation, is provided.",
keywords = "Distributional lexical semantics, Kernel methods, Question classification",
author = "Danilo Croce and Simone Filice and Roberto Basili",
year = "2015",
doi = "10.1007/978-3-319-14206-7_6",
language = "English",
volume = "589",
pages = "115--134",
journal = "Studies in Computational Intelligence",
issn = "1860-949X",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Distributional models for lexical semantics

T2 - An investigation of different representations for natural language learning

AU - Croce, Danilo

AU - Filice, Simone

AU - Basili, Roberto

PY - 2015

Y1 - 2015

N2 - Language learning systems usually generalize linguistic observations into rules and patterns that are statistical models of higher level semantic inferences. When the availability of training data is scarce, lexical information can be limited by data sparseness effects and generalization is thus needed. Distributional models represent lexical semantic information in terms of the basic co-occurrences between words in large-scale text collections. As recent works already address, the definition of proper distributional models as well as methods able to express the meaning of phrases or sentences as operations on lexical representations is a complex problem, and a still largely open issue. In this paper, a perspective centered on Convolution Kernels is discussed and the formulation of a Partial Tree Kernel that integrates syntactic information and lexical generalization is studied. Moreover a large scale investigation of different representation spaces, each capturing a different linguistic relation, is provided.

AB - Language learning systems usually generalize linguistic observations into rules and patterns that are statistical models of higher level semantic inferences. When the availability of training data is scarce, lexical information can be limited by data sparseness effects and generalization is thus needed. Distributional models represent lexical semantic information in terms of the basic co-occurrences between words in large-scale text collections. As recent works already address, the definition of proper distributional models as well as methods able to express the meaning of phrases or sentences as operations on lexical representations is a complex problem, and a still largely open issue. In this paper, a perspective centered on Convolution Kernels is discussed and the formulation of a Partial Tree Kernel that integrates syntactic information and lexical generalization is studied. Moreover a large scale investigation of different representation spaces, each capturing a different linguistic relation, is provided.

KW - Distributional lexical semantics

KW - Kernel methods

KW - Question classification

UR - http://www.scopus.com/inward/record.url?scp=84926639516&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84926639516&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-14206-7_6

DO - 10.1007/978-3-319-14206-7_6

M3 - Article

AN - SCOPUS:84926639516

VL - 589

SP - 115

EP - 134

JO - Studies in Computational Intelligence

JF - Studies in Computational Intelligence

SN - 1860-949X

ER -