Extractive Summarization with Discourse Graphs

This program is implemented base on Windows 10 x64 System. Using Python 3.8 and PyTorch with CUDA 12.1. Update to newer version of PyTorch using instructions from PyTorch Install Guide

Make sure you have the required dependencies installed by running:

pip install -r requirements.txt

• --train_path: Specify the path to the training data. (Default: None)
• --test_path: Specify the path to the test data. (Default: None)
• --labels_path: Specify the path to the labels. (Default: None)
• --model: Specify the model to use. (Default: DecisionTree)
• --result_path: Specify the path to save the result. (Default: test_labels.json)

Use the following command to run the main script:

python main.py --train_path /path/to/training_data --test_path /path/to/test_data --labels_path /path/to/labels --model YourModel --result_path /path/to/result_file.json

Replace /path/to/training_data, /path/to/test_data, /path/to/labels, YourModel, and /path/to/result_file.json with your actual paths and model choice.

In the end, convert the results in json format to kaggle-compatible submission.csv with:

python make_submission.py --json_path /path/to/result_file.json

Replace /path/to/result_file.json with your actual paths.

• DecisionTree
• RandomForest
• XGBoost
• TwoModels
• LSTM
• GCN
• GAT
• GraphSAGE