https://github.com/sriramsk1999/tda-coherence-modeling

https://github.com/sriramsk1999/tda-coherence-modeling

Last synced: 1 day ago
JSON representation

• Host: GitHub
• URL: https://github.com/sriramsk1999/tda-coherence-modeling
• Owner: sriramsk1999
• License: apache-2.0
• Created: 2023-02-18T07:11:47.000Z (over 1 year ago)
• Default Branch: master
• Last Pushed: 2024-04-02T13:40:23.000Z (8 months ago)
• Last Synced: 2024-04-02T14:40:47.878Z (8 months ago)
• Language: Python
• Size: 63.5 KB
• Stars: 0
• Watchers: 1
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Coherence Modeling with TDA

[**Beyond Words: A Topological Exploration of Coherence in Text Documents**](https://openreview.net/forum?id=QJxVhljAyS)

The Second Tiny Papers Track at ICLR 2024

## Setup

```py

conda env create -f environment.yml

conda activate tda-modeling-env

```

## Usage

Computing TDA features for a dataset:

```py

python feature_gen.py --cuda 0 --data_name clinton_train --input_dir GCDC_Dataset/ --output_dir gcdc_tda_features --batch_size 100

```

Train/test MLP using generated TDA features:

```py

python predict_tda.py --input_dir GCDC_Dataset/ --feat_dir gcdc_tda_features/  --domain clinton

```

## Data

- **GCDC** - Refer [GCDC-Corpus](https://github.com/aylai/GCDC-corpus) for the source

## Acknowledgements

We thank the authors of [Artificial Text Detection via Examining the Topology of Attention Maps (EMNLP 2021)](https://github.com/danchern97/tda4atd) for publishing their code. 

## Citing

If you find this code useful for your research, please cite the following paper:

```bibtex

@inproceedings{jain2024beyond,

  title={Beyond Words: A Topological Exploration of Coherence in Text Documents},

  author={Jain, Samyak and Singhal, Rishi and Krishna, Sriram and Singla, Yaman K and Shah, Rajiv Ratn},

  booktitle={The Second Tiny Papers Track at ICLR 2024}

}

```