Unsupervised measure of word similarity: How to outperform co-occurrence and vector cosine in VSMs

Enrico Santus; Tin Shing Chiu; Qin Lu; Alessandro Lenci; Chu-ren Huang

Unsupervised measure of word similarity: How to outperform co-occurrence and vector cosine in VSMs

Enrico Santus, Tin Shing Chiu, Qin Lu, Alessandro Lenci, Chu-ren Huang

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

10 Citations (Scopus)

Abstract

In this paper, we claim that vector cosine - which is generally considered among the most efficient unsupervised measures for identifying word similarity in Vector Space Models - can be outperformed by an unsupervised measure that calculates the extent of the intersection among the most mutually dependent contexts of the target words. To prove it, we describe and evaluate APSyn, a variant of the Average Precision that, without any optimization, outperforms the vector cosine and the co-occurrence on the standard ESL test set, with an improvement ranging between +9.00% and +17.98%, depending on the number of chosen top contexts.

Original language	English
Title of host publication	30th AAAI Conference on Artificial Intelligence, AAAI 2016
Publisher	AAAI press
Pages	4260-4261
Number of pages	2
ISBN (Electronic)	9781577357605
Publication status	Published - 1 Jan 2016
Event	30th AAAI Conference on Artificial Intelligence, AAAI 2016 - Phoenix Convention Center, Phoenix, United States Duration: 12 Feb 2016 → 17 Feb 2016

Conference

Conference	30th AAAI Conference on Artificial Intelligence, AAAI 2016
Country/Territory	United States
City	Phoenix
Period	12/02/16 → 17/02/16

ASJC Scopus subject areas

Artificial Intelligence

Cite this

@inproceedings{1df8f9d8ef0948b2a9e72e03565319d5,

title = "Unsupervised measure of word similarity: How to outperform co-occurrence and vector cosine in VSMs",

abstract = "In this paper, we claim that vector cosine - which is generally considered among the most efficient unsupervised measures for identifying word similarity in Vector Space Models - can be outperformed by an unsupervised measure that calculates the extent of the intersection among the most mutually dependent contexts of the target words. To prove it, we describe and evaluate APSyn, a variant of the Average Precision that, without any optimization, outperforms the vector cosine and the co-occurrence on the standard ESL test set, with an improvement ranging between +9.00% and +17.98%, depending on the number of chosen top contexts.",

author = "Enrico Santus and Chiu, {Tin Shing} and Qin Lu and Alessandro Lenci and Chu-ren Huang",

year = "2016",

month = jan,

day = "1",

language = "English",

pages = "4260--4261",

booktitle = "30th AAAI Conference on Artificial Intelligence, AAAI 2016",

publisher = "AAAI press",

note = "30th AAAI Conference on Artificial Intelligence, AAAI 2016 ; Conference date: 12-02-2016 Through 17-02-2016",

}

Unsupervised measure of word similarity: How to outperform co-occurrence and vector cosine in VSMs. / Santus, Enrico; Chiu, Tin Shing; Lu, Qin et al.
30th AAAI Conference on Artificial Intelligence, AAAI 2016. AAAI press, 2016. p. 4260-4261.

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

TY - GEN

T1 - Unsupervised measure of word similarity: How to outperform co-occurrence and vector cosine in VSMs

AU - Santus, Enrico

AU - Chiu, Tin Shing

AU - Lu, Qin

AU - Lenci, Alessandro

AU - Huang, Chu-ren

PY - 2016/1/1

Y1 - 2016/1/1

N2 - In this paper, we claim that vector cosine - which is generally considered among the most efficient unsupervised measures for identifying word similarity in Vector Space Models - can be outperformed by an unsupervised measure that calculates the extent of the intersection among the most mutually dependent contexts of the target words. To prove it, we describe and evaluate APSyn, a variant of the Average Precision that, without any optimization, outperforms the vector cosine and the co-occurrence on the standard ESL test set, with an improvement ranging between +9.00% and +17.98%, depending on the number of chosen top contexts.

AB - In this paper, we claim that vector cosine - which is generally considered among the most efficient unsupervised measures for identifying word similarity in Vector Space Models - can be outperformed by an unsupervised measure that calculates the extent of the intersection among the most mutually dependent contexts of the target words. To prove it, we describe and evaluate APSyn, a variant of the Average Precision that, without any optimization, outperforms the vector cosine and the co-occurrence on the standard ESL test set, with an improvement ranging between +9.00% and +17.98%, depending on the number of chosen top contexts.

UR - http://www.scopus.com/inward/record.url?scp=85007200179&partnerID=8YFLogxK

M3 - Conference article published in proceeding or book

SP - 4260

EP - 4261

BT - 30th AAAI Conference on Artificial Intelligence, AAAI 2016

PB - AAAI press

T2 - 30th AAAI Conference on Artificial Intelligence, AAAI 2016

Y2 - 12 February 2016 through 17 February 2016

ER -

Unsupervised measure of word similarity: How to outperform co-occurrence and vector cosine in VSMs

Abstract

Conference

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this