Awesome

SemEval-2024 Task 3: Codalab Service for THOR-ECAC •

Update 05 March 2024: The quick arXiv paper breakdowns 🔨 are @ Twitter/X post

This repository shares data and submission-related code for training and handling results of
THoR-ECAC framework, as a part of the SemEval-2024 paper nicolay-r at SemEval-2024 Task 3: Using Flan-T5 for Reasoning Emotion Cause in Conversations with Chain-of-Thought on Emotion States

👉THoR-ECAC framework👈

Usage

1. Install necessary project dependencies as follows:

pip install -r dependencies.txt

2. Use download.py scripts for fetching ⬇️ all the task-related resources:

python download.py

3. Use the following shared scripts:

• Resources Preparation for pair-based experiments in context:
  • 0_emotion_state.py -- script for pretraining data preparation, based on states;
  • 0_emotion_cause.py -- script for fine-tuning data preparation, based on causes;
  • 1_ps_vocab.py -- vocabulary preparation for manual spans corrections.
• Spans correction algorithm implementation of the vocabulary-based spans-correction technique mentioned in paper.
• Codalab submissions forming:
  • 2_submit -- script for forming *.json.zip archive, compatible for submitting on Codalab platfom.
• Conversation data analysis:
  • task_statistics_json.py -- json data analyzer;
  • task_statistics_submission.py -- *json.zip submissions analyzer.

References

You can cite this work or THoR-ECAC framework as follows:

@article{rusnachenko2024nicolayr,
  title={nicolay-r at SemEval-2024 Task 3: Using Flan-T5 for Reasoning Emotion Cause in Conversations with Chain-of-Thought on Emotion States},
  booktitle = "Proceedings of the Annual Conference of the North American Chapter of the Association for Computational Linguistics",
  author={Nicolay Rusnachenko and Huizhi Liang},
  year= "2024",
  month= jun,
  address = "Mexico City, Mexico",
  publisher = "Association for Computational Linguistics"
}