google-research

This repository contains reference code for the ICLR-23 paper "KwikBucks: Correlation Clustering with Cheap-Weak and Expensive-Strong signals". The code implements the main KwikBucks algorithm of the paper (it is implemented as qwick_cluster_using_ordering in model_utils.py) as well as the baselines stated in the experimental section of the paper.

Datasets

The datasets can be downlowded from here: https://storage.googleapis.com/gresearch/kwikbucks/kwikbucks.zip

Installation

• Download the code.
• Download the the dataset and put it under kwikbucks/data/.
• Run run.sh to install requirements.txt and run the code.

Citation

If you use the code, please cite our paper.

Silwal, S., Ahmadian, S., Nystrom, A., McCallum, A., Ramachandran, D., Kazemi, M., "KwikBucks: Correlation Clustering with Cheap-Weak and Expensive-Strong Signals", The Eleventh International Conference on Learning Representations (ICLR), 2023.

@inproceedings{silwal2023kwikbucks,
  title={KwikBucks: Correlation Clustering with Cheap-Weak and Expensive-Strong Signals},
  author={Silwal, Sandeep and Ahmadian, Sara and Nystrom, Andrew and McCallum, Andrew and Ramachandran, Deepak and Kazemi, Seyed Mehran},
  booktitle={The Eleventh International Conference on Learning Representations (ICLR)},
  year={2023}
}

Contact us

For questions or comments about the implementation, please contact mehrankazemi@google.com or silwal@mit.edu.

Disclaimer

This is not an official Google product.