NewSoMe Corpus of Opinion in News Reports

Item Name: NewSoMe Corpus of Opinion in News Reports
Author(s): Roser Sauri, Judith Domingo, Toni Badia
LDC Catalog No.: LDC2015T17
ISBN: 1-58563-731-9
ISLRN: 793-803-205-712-6
Release Date: September 15, 2015
Member Year(s): 2015
DCMI Type(s): Text
Data Source(s): newswire, web collection
Application(s): discourse analysis, information extraction
Language(s): Spanish, Catalan, Portuguese
Language ID(s): spa, cat, por
License(s): LDC User Agreement for Non-Members
Online Documentation: LDC2015T17 Documents
Licensing Instructions: Subscription & Standard Members, and Non-Members
Citation: Sauri, Roser, Judith Domingo, and Toni Badia. NewSoMe Corpus of Opinion in News Reports LDC2015T17. Web Download. Philadelphia: Linguistic Data Consortium, 2015.

Introduction

NewSoMe Corpus of Opinion in News Reports was compiled at Barcelona Media and consists of Spanish, Catalan and Portuguese news reports annotated for opinions. It is part of the NewSoMe (News and Social Media) set of corpora presenting opinion annotations across several genres and covering multiple languages. NewSoMe is the result of an effort to build a unifying annotation framework for analyzing opinion in different genres, ranging from controlled text, such as news reports, to diverse types of user-generated content that includes blogs, product reviews and microblogs.

Data

The source data in this release was obtained from various newspaper websites and consists of approximately 200 documents in each of Spanish, Catalan and Portuguese. The annotation was carried out manually through the crowdsourcing platform CrowdFlower with seven annotations per layer that were aggregated for this data set. The layers annotated were topic, segment, cue, subjectivity, polarity and intensity.

Data is presented as UTF-8 either as plain text or in CSV files.

Samples

Please view the following samples.

Updates

None at this time.

Available Media

View Fees





Login for the applicable fee