Please use this identifier to cite or link to this item:

Title: How does irony affect sentiment analysis tools?
Authors: Weitzel, Leila
Freire, Raul
Quaresma, Paulo
Gonçalves, Teresa
Editors: Pereira, Francisco
Machado, Penousal
Costa, Ernesto
Cardoso, Amilcar
Issue Date: 2015
Publisher: Springer International Publishing
Abstract: Sentiment analysis applications have spread to many domains: from consumer products, healthcare and financial services to political elections and social events. A common task in opinion mining is to classify an opinionated document into a positive or negative opinion. In this paper, a study of different methodologies is conducted to rank polarity as to better know how the ironic messages affect sentiment analysis tools. The study provides an initial under- standing of how irony affects the polarity detection. From the statistic point of view, we realize that there are no significant differences between methodolo- gies. To better understand the phenomenon, it is essential to apply different methods, such as SentiWordNet, based on Lexicon. In this sense, as future work, we aim to explore the use of Lexicon based tools, thus measuring and comparing the attained results.
Type: article
Appears in Collections:INF - Publicações - Artigos em Revistas Internacionais Com Arbitragem Científica

Files in This Item:

File Description SizeFormat
GAIA- EPIA-2015v2.pdf192.7 kBAdobe PDFView/OpenRestrict Access. You can Request a copy!
FacebookTwitterDeliciousLinkedInDiggGoogle BookmarksMySpaceOrkut
Formato BibTex mendeley Endnote Logotipo do DeGóis 

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.


Dspace Dspace
DSpace Software, version 1.6.2 Copyright © 2002-2008 MIT and Hewlett-Packard - Feedback
UEvora B-On Curriculum DeGois