Accounting for Negation in Sentiment Analysis

BMWT Heerschop; P van Iterson; Alexander Hogenboom; Flavius Frasincar; U Kaymak

Accounting for Negation in Sentiment Analysis

BMWT Heerschop, P van Iterson, Alexander Hogenboom, Flavius Frasincar, U Kaymak

Research output: Contribution to conference › Poster › Academic

Abstract

Automated ways of analyzing sentiment in Web data are becoming more and more urgent as virtual utterances of opinions or sentiment are becoming increasingly abundant on the Web. The role of negation in sentiment analysis has been explored only to a limited extent until now. In this paper, we investigate the impact of accounting for negation in sentiment analysis. To this end, we utilize a basic sentiment analysis framework – consisting of a wordbank creation part and a document scoring part – taking into account negation. Our experimental results show that by accounting for negation, precision relative to human ratings increases with 1.17%. On a subset of selected documents containing negated words, precision increases with 2.23%.

Original language	English
Pages	38-39
Number of pages	2
Publication status	Published - 4 Feb 2011
Event	Eleventh Dutch-Belgian Information Retrieval Workshop (DIR 2011) - Amsterdam, The Netherlands Duration: 4 Feb 2011 → 4 Feb 2011

Conference

Conference	Eleventh Dutch-Belgian Information Retrieval Workshop (DIR 2011)
City	Amsterdam, The Netherlands
Period	4/02/11 → 4/02/11

Cite this

@conference{95fddd9ce1ca4f2fa60309d6a9c7a614,

title = "Accounting for Negation in Sentiment Analysis",

abstract = "Automated ways of analyzing sentiment in Web data are becoming more and more urgent as virtual utterances of opinions or sentiment are becoming increasingly abundant on the Web. The role of negation in sentiment analysis has been explored only to a limited extent until now. In this paper, we investigate the impact of accounting for negation in sentiment analysis. To this end, we utilize a basic sentiment analysis framework – consisting of a wordbank creation part and a document scoring part – taking into account negation. Our experimental results show that by accounting for negation, precision relative to human ratings increases with 1.17%. On a subset of selected documents containing negated words, precision increases with 2.23%.",

author = "BMWT Heerschop and {van Iterson}, P and Alexander Hogenboom and Flavius Frasincar and U Kaymak",

year = "2011",

month = feb,

day = "4",

language = "English",

pages = "38--39",

note = "Eleventh Dutch-Belgian Information Retrieval Workshop (DIR 2011) ; Conference date: 04-02-2011 Through 04-02-2011",

}

TY - CONF

T1 - Accounting for Negation in Sentiment Analysis

AU - Heerschop, BMWT

AU - van Iterson, P

AU - Hogenboom, Alexander

AU - Frasincar, Flavius

AU - Kaymak, U

PY - 2011/2/4

Y1 - 2011/2/4

N2 - Automated ways of analyzing sentiment in Web data are becoming more and more urgent as virtual utterances of opinions or sentiment are becoming increasingly abundant on the Web. The role of negation in sentiment analysis has been explored only to a limited extent until now. In this paper, we investigate the impact of accounting for negation in sentiment analysis. To this end, we utilize a basic sentiment analysis framework – consisting of a wordbank creation part and a document scoring part – taking into account negation. Our experimental results show that by accounting for negation, precision relative to human ratings increases with 1.17%. On a subset of selected documents containing negated words, precision increases with 2.23%.

AB - Automated ways of analyzing sentiment in Web data are becoming more and more urgent as virtual utterances of opinions or sentiment are becoming increasingly abundant on the Web. The role of negation in sentiment analysis has been explored only to a limited extent until now. In this paper, we investigate the impact of accounting for negation in sentiment analysis. To this end, we utilize a basic sentiment analysis framework – consisting of a wordbank creation part and a document scoring part – taking into account negation. Our experimental results show that by accounting for negation, precision relative to human ratings increases with 1.17%. On a subset of selected documents containing negated words, precision increases with 2.23%.

UR - http://ilps.science.uva.nl/dir2011/DIR-2011-proceedings.pdf

M3 - Poster

SP - 38

EP - 39

T2 - Eleventh Dutch-Belgian Information Retrieval Workshop (DIR 2011)

Y2 - 4 February 2011 through 4 February 2011

ER -

Accounting for Negation in Sentiment Analysis

Abstract

Conference

Other files and links

Fingerprint

Cite this