MuTEC: An end to end Multi-Task learning framework for extractingEmotions and emotion Cause in conversations

Top-level structure

.
├── MuTEC_CSE                 # Task1: Cause Span Extraction End2End Architecture
├── MuTEC_CEE                 # Task2: Causal Emotion Entailment End2End Architecture
├── two-step                  # Task1: Cause Span Extraction Two Step Architecture
└── README.md

Inside every directory you can find a README that explains the structure of that directory.

Dataset is taken from RECCON paper: https://github.com/declare-lab/RECCON

MuTEC balanced dataset: https://drive.google.com/drive/folders/10Vn-RWEGODxaHd1vD9mKc26iB1qe-o48?usp=sharing

License

The code follows CC-BY-NC license. Thus, users can share and adapt the code if they give credit to us and do not use for any commercial purposes.

Paper Link

https://arxiv.org/abs/2211.03742