Sentiment Analysis for Emotional Speech Synthesis in a News Dialogue System

Hiroaki Takatsu, Ryota Ando, Yoichi Matsuyama, Tetsunori Kobayashi


Abstract
As smart speakers and conversational robots become ubiquitous, the demand for expressive speech synthesis has increased. In this paper, to control the emotional parameters of the speech synthesis according to certain dialogue contents, we construct a news dataset with emotion labels (“positive,” “negative,” or “neutral”) annotated for each sentence. We then propose a method to identify emotion labels using a model combining BERT and BiLSTM-CRF, and evaluate its effectiveness using the constructed dataset. The results showed that the classification model performance can be efficiently improved by preferentially annotating news articles with low confidence in the human-in-the-loop machine learning framework.
Anthology ID:
2020.coling-main.440
Volume:
Proceedings of the 28th International Conference on Computational Linguistics
Month:
December
Year:
2020
Address:
Barcelona, Spain (Online)
Editors:
Donia Scott, Nuria Bel, Chengqing Zong
Venue:
COLING
SIG:
Publisher:
International Committee on Computational Linguistics
Note:
Pages:
5013–5025
Language:
URL:
https://aclanthology.org/2020.coling-main.440
DOI:
10.18653/v1/2020.coling-main.440
Bibkey:
Cite (ACL):
Hiroaki Takatsu, Ryota Ando, Yoichi Matsuyama, and Tetsunori Kobayashi. 2020. Sentiment Analysis for Emotional Speech Synthesis in a News Dialogue System. In Proceedings of the 28th International Conference on Computational Linguistics, pages 5013–5025, Barcelona, Spain (Online). International Committee on Computational Linguistics.
Cite (Informal):
Sentiment Analysis for Emotional Speech Synthesis in a News Dialogue System (Takatsu et al., COLING 2020)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-4/2020.coling-main.440.pdf