EmotionX-AR: CNN-DCNN autoencoder based Emotion Classifier

Sopan Khosla


Abstract
In this paper, we model emotions in EmotionLines dataset using a convolutional-deconvolutional autoencoder (CNN-DCNN) framework. We show that adding a joint reconstruction loss improves performance. Quantitative evaluation with jointly trained network, augmented with linguistic features, reports best accuracies for emotion prediction; namely joy, sadness, anger, and neutral emotion in text.
Anthology ID:
W18-3507
Volume:
Proceedings of the Sixth International Workshop on Natural Language Processing for Social Media
Month:
July
Year:
2018
Address:
Melbourne, Australia
Editors:
Lun-Wei Ku, Cheng-Te Li
Venue:
SocialNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
37–44
Language:
URL:
https://aclanthology.org/W18-3507
DOI:
10.18653/v1/W18-3507
Bibkey:
Cite (ACL):
Sopan Khosla. 2018. EmotionX-AR: CNN-DCNN autoencoder based Emotion Classifier. In Proceedings of the Sixth International Workshop on Natural Language Processing for Social Media, pages 37–44, Melbourne, Australia. Association for Computational Linguistics.
Cite (Informal):
EmotionX-AR: CNN-DCNN autoencoder based Emotion Classifier (Khosla, SocialNLP 2018)
Copy Citation:
PDF:
https://preview.aclanthology.org/ml4al-ingestion/W18-3507.pdf