2.50
Hdl Handle:
http://hdl.handle.net/10547/275675
Title:
Determining the polarity of postings for discussion search
Authors:
Frommholz, Ingo ( 0000-0002-5622-5132 ) ; Lechtenfeld, Marc
Abstract:
When performing discussion search it might be desirable to consider non-topical measures like the number of positive and negative replies to a posting, for instance as one possible indicator for the trustworthiness of a comment. Systems like POLAR are able to integrate such values into the retrieval function. To automatically detect the polarity of postings, they need to be classified into positive and negative ones w.r.t.\ the comment or document they are annotating. We present a machine learning approach for polarity detection which is based on Support Vector Machines. We discuss and identify appropriate term and context features. Experiments with ZDNet News show that an accuracy of around 79\%-80\% can be achieved for automatically classifying comments according to their polarity.
Affiliation:
University of Duisburg-Essen, Germany
Citation:
Frommholz, I. and Lechtenfeld, M. (2008) 'Determining the Polarity of Postings for Discussion Search,' in Proc. of the ``Information Retrieval 2008’' Workshop at LWA 2008
Publisher:
Gesellschaft für Informatik e.V.
Issue Date:
2008
URI:
http://hdl.handle.net/10547/275675
Additional Links:
http://www.is.inf.uni-due.de/bib/docs/Frommholz_Lechtenfeld_08.html
Type:
Conference papers, meetings and proceedings
Language:
en
Appears in Collections:
Centre for Research in Distributed Technologies (CREDIT)

Full metadata record

DC FieldValue Language
dc.contributor.authorFrommholz, Ingoen_GB
dc.contributor.authorLechtenfeld, Marcen_GB
dc.date.accessioned2013-03-22T12:12:03Z-
dc.date.available2013-03-22T12:12:03Z-
dc.date.issued2008-
dc.identifier.citationFrommholz, I. and Lechtenfeld, M. (2008) 'Determining the Polarity of Postings for Discussion Search,' in Proc. of the ``Information Retrieval 2008’' Workshop at LWA 2008en_GB
dc.identifier.urihttp://hdl.handle.net/10547/275675-
dc.description.abstractWhen performing discussion search it might be desirable to consider non-topical measures like the number of positive and negative replies to a posting, for instance as one possible indicator for the trustworthiness of a comment. Systems like POLAR are able to integrate such values into the retrieval function. To automatically detect the polarity of postings, they need to be classified into positive and negative ones w.r.t.\ the comment or document they are annotating. We present a machine learning approach for polarity detection which is based on Support Vector Machines. We discuss and identify appropriate term and context features. Experiments with ZDNet News show that an accuracy of around 79\%-80\% can be achieved for automatically classifying comments according to their polarity.en_GB
dc.language.isoenen
dc.publisherGesellschaft für Informatik e.V.en_GB
dc.relation.urlhttp://www.is.inf.uni-due.de/bib/docs/Frommholz_Lechtenfeld_08.htmlen_GB
dc.subjectdiscussion searchen_GB
dc.subjectSupport Vector Machinesen_GB
dc.subjectinformation retrievalen_GB
dc.subjectpolarity detectionen_GB
dc.titleDetermining the polarity of postings for discussion searchen
dc.typeConference papers, meetings and proceedingsen
dc.contributor.departmentUniversity of Duisburg-Essen, Germanyen_GB
This item is licensed under a Creative Commons License
Creative Commons
All Items in UOBREP are protected by copyright, with all rights reserved, unless otherwise indicated.