EmotionLines: An Emotion Corpus of Multi-Party Conversations
Paper • 1802.08379 • Published
Datasets:
Dataset Viewer
The dataset viewer is not available for this dataset.
Unexpected token '<', "<html>
<h"... is not valid JSON
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
MELD Preprocessed for SER
This dataset is the manually preprocessed audio only version of MELD, only audio IDs, utterance transcriptions, dialogue IDs and Utterance IDs were extracted.
S. Poria, D. Hazarika, N. Majumder, G. Naik, R. Mihalcea,
E. Cambria. MELD: A Multimodal Multi-Party Dataset
for Emotion Recognition in Conversation. (2018)
Chen, S.Y., Hsu, C.C., Kuo, C.C. and Ku, L.W.
EmotionLines: An Emotion Corpus of Multi-Party
Conversations. arXiv preprint arXiv:1802.08379 (2018).
- Downloads last month
- 20