ezCoref: Towards unifying annotation guidelines for coreference resolution

This is the official repository for our paper, ezCoref: Towards unifying annotation guidelines for coreference resolution. ezCoref is an annotation methodology for coreference resolution, comprising of a crowdsourcing-friendly annotation tool and an interative tutorial. ezCoref has been validated on Amazon Mechanical Turk.

This repository contains code for launching ezCoref (Tool and Tutorial) and crowdsourced annotations collected via ezCoref.

Try out a demo of ezCoref tutorial and annotating a portion of "Alice in Wonderland" using ezCoref.

Update (May 2023): Code for tool and annotated data are now available.

Update (Feb 2023): Demo of the tool and tutorial are now available.

Citation Information

If you use ezCoref, please cite it as follows:

@inproceedings{gupta-etal-2023-ezcoref,
    title = "ez{C}oref: Towards Unifying Annotation Guidelines for Coreference Resolution",
    author = "Gupta, Ankita  and Karpinska, Marzena  and Zhao, Wenlong  and Krishna, Kalpesh  and Merullo, Jack  and Yeh, Luke  and Iyyer, Mohit  and O{'}Connor, Brendan",
    booktitle = "Findings of the Association for Computational Linguistics: EACL 2023",
    year = "2023",
    address = "Dubrovnik, Croatia",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2023.findings-eacl.24",
    pages = "312--330",
}