Introduction

This repository was used in our paper:

“Transition-based Directed Graph Construction for Emotion-Cause Pair Extraction”
Chuang Fan, Chaofa Yuan, Jiachen Du, Lin Gui, Min Yang, Ruifeng Xu. ACL 2020

Please cite our paper if you use this code.

Prerequisites

Python 3.6
Pytorch 1.1.0
CUDA 10.0
BERT - Our bert model is adapted from this implementation: https://github.com/huggingface/pytorch-pretrained-BERT

Descriptions

Data - A dir where contains resources used in this code.

• bert-base-chinese: Put the download Pytorch bert model here.
• DataSplits: A dir where contains 20 different training/validation/test splits in a ratio of 8:1:1. Each sub-dir contains four file: saved_results.txt, train.pkl, valid.pkl and test.pkl.
  • saved_results.txt: The results of test set for emotion extraction, cause extraction and emotion-cause pair extraction. We adopt early stopping strategy, and the highest F-measure model on the validation set is used to evaluate the test set.
  • train.pkl: A list where contains two items. train[0] is a list of document and train[1] is a list of the correspondding emotion-cause pairs. For example, train[0][0]="Last week, I lost my phone where shopping, I feel sad now", then train[1][0]=[(2, 1)].
  • valid.pkl: Similar to train.pkl.
  • test.pkl: Similar to train.pkl.
• doc2pair.pkl: A dict where the key is the content of a document, and the value is the correspondding emotion-cause pairs.

Utils - A dir where contains several python scripts used in this code.

• Evaluation.py: Used to evaluate the performance of the proposed model.
• Metrics.py: Metrics for emotion extraction, cause extraction and emotion-cause pair extractions.
• PrepareData.py: The scipt for preparing data.
• Transform.py: Transforming documents to a sequence of defined actions and parser states from left-to-right based on the emotion-cause pairs.

Config.py - The script holds all the model configuration.
TransModule.py - The script where contains the proposed transition-based model.
Run.py - The main script to train and evaluate the proposed transition-based model on different splits.

Usage

python3 Run.py