Similarity of Semantic Relations

There are at least two kinds of similarity. is correspondence between relations, in contrast with , which is correspondence between attributes. When two words have a high degree of attributional similarity, we call them . When two pairs of words have a high degree of relational similarity, we say th...

Full description

Saved in:
Bibliographic Details
Published in:Computational linguistics - Association for Computational Linguistics 2006-09, Vol.32 (3), p.379-416
Main Author: Turney, Peter D.
Format: Article
Language:English
Subjects:
Applied linguistics
Computational linguistics
Linguistics
Semantics
Statistical analysis
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:There are at least two kinds of similarity. is correspondence between relations, in contrast with , which is correspondence between attributes. When two words have a high degree of attributional similarity, we call them . When two pairs of words have a high degree of relational similarity, we say that their relations are . For example, the word pair mason:stone is analogous to the pair carpenter:wood. This article introduces Latent Relational Analysis (LRA), a method for measuring relational similarity. LRA has potential applications in many areas, including information extraction, word sense disambiguation, and information retrieval. Recently the Vector Space Model (VSM) of information retrieval has been adapted to measuring relational similarity, achieving a score of 47% on a collection of 374 college-level multiple-choice word analogy questions. In the VSM approach, the relation between a pair of words is characterized by a vector of frequencies of predefined patterns in a large corpus. LRA extends the VSM approach in three ways: (1) The patterns are derived automatically from the corpus, (2) the Singular Value Decomposition (SVD) is used to smooth the frequency data, and (3) automatically generated synonyms are used to explore variations of the word pairs. LRA achieves 56% on the 374 analogy questions, statistically equivalent to the average human score of 57%. On the related problem of classifying semantic relations, LRA achieves similar gains over the VSM.
ISSN:0891-2017
1530-9312
DOI:10.1162/coli.2006.32.3.379