Detecting speculations, contrasts and conditionals in consumer reviews
Författare
Redaktör
- Balahur Alexandra
- van der Goot Erik
- Vossen Piek
- Montoyo Andrés
Summary, in English
A support vector classifier was compared to a lexicon-based approach for the task of detecting the stance categories speculation, contrast and conditional in English consumer reviews. Around 3,000 training instances were required to achieve a stable performance of an F-score of 90 for speculation. This outperformed the lexicon-based approach, for which an Fscore of just above 80 was achieved. The machine learning results for the other two categories showed a lower average (an approximate F-score of 60 for contrast and 70 for conditional), as well as a larger variance, and were only slightly better than lexicon matching. Therefore, while machine learning was successful for detecting speculation, a well-curated lexicon might be a more suitable approach for detecting contrast and conditional.
Avdelning/ar
- Engelska
- Language, Cognition and Discourse@Lund (LCD@L)
Publiceringsår
2015
Språk
Engelska
Sidor
162-168
Publikation/Tidskrift/Serie
6th Workshop on Computational Approaches to Subjectivity, Sentiment & Social Media Analysis, WASSA 2015 : Workshop proceedings
Länkar
Dokumenttyp
Konferensbidrag
Förlag
Association for Computational Linguistics
Ämne
- Languages and Literature
- Computer Science
Nyckelord
- consumer reviews
- support vector classifier
- stance
Conference name
6th Workshop on Computational Approaches to Subjectivity, Sentiment & Social Media Analysis (WASSA '15)
Conference date
2015-09-17
Status
Published
Projekt
- StaViCTA - Advances in the description and explanation of stance in discourse using visual and computational text analytics
Forskningsgrupp
- Language, Cognition and Discourse@Lund (LCD@L)
ISBN/ISSN/Övrigt
- ISBN: 978-1-941643-32-7