EmotiKLUE at IEST 2018: Topic-Informed Classification of Implicit Emotions

Beitrag bei einer Tagung
(Originalarbeit)


Details zur Publikation

Autor(en): Proisl T, Heinrich P, Kabashi B, Evert S
Herausgeber: Balahur A, Mohammad SM, Hoste V, Klinger R
Verlag: Association for Computational Linguistics
Verlagsort: Brussels
Jahr der Veröffentlichung: 2018
Tagungsband: Proceedings of the 9th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis
Seitenbereich: 235–242
Sprache: Englisch


Abstract

EmotiKLUE is a submission to the Implicit Emotion Shared Task. It is a deep learning system that combines independent representations of the left and right contexts of the emotion word with the topic distribution of an LDA topic model. EmotiKLUE achieves a macro average F 1 score of 67.13%, significantly outperforming the baseline produced by a simple ML classifier. Further enhancements after the evaluation period lead to an improved F 1 score of 68.10%


FAU-Autoren / FAU-Herausgeber

Evert, Stefan Prof. Dr.
Lehrstuhl für Korpus- und Computerlinguistik
Heinrich, Philipp
Lehrstuhl für Korpus- und Computerlinguistik
Kabashi, Besim Dr.
Lehrstuhl für Korpus- und Computerlinguistik
Proisl, Thomas
Lehrstuhl für Korpus- und Computerlinguistik


Zitierweisen

APA:
Proisl, T., Heinrich, P., Kabashi, B., & Evert, S. (2018). EmotiKLUE at IEST 2018: Topic-Informed Classification of Implicit Emotions. In Balahur A, Mohammad SM, Hoste V, Klinger R (Eds.), Proceedings of the 9th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis (pp. 235–242). Brüssel, BE: Brussels: Association for Computational Linguistics.

MLA:
Proisl, Thomas, et al. "EmotiKLUE at IEST 2018: Topic-Informed Classification of Implicit Emotions." Proceedings of the 9th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis, Brüssel Ed. Balahur A, Mohammad SM, Hoste V, Klinger R, Brussels: Association for Computational Linguistics, 2018. 235–242.

BibTeX: 

Zuletzt aktualisiert 2018-13-11 um 09:38