OntoGUM: Evaluating Contextualized SOTA Coreference Resolution on 12 More Genres

Introduction

This repository contains the code for building up the OntoGUM dataset from:

Prerequisites

Python >= 3.6
Download GUM from https://github.com/amir-zeldes/gum and put the folder in the home directory of this repo

Rebuilding the dataset

You can either use the scripts in this repository or the built bot from GUM to rebuild the dataset.

To rebuild the dataset using this repo:

Run main.py to start the conversion after following the prerequisites.
Adjust the arguments in main.py to output different formats. Please note that if you want to test models trained on OntoNotes, the conll format is needed.

To rebuild the dataset from GUM:

Also Check here for differences between GUM and OntoNotes schema.

Follow the instructions in the GUM repo to build up the dataset (including reddit data)
Find the OntoGUM data (tsv and conll) under /gum/_build/target/coref/ontogum

Output

Two output formats are currently supported: tsv and conll. The default output is tsv. If you would like to have the conll format, specify it with the argument --out_format

Visualization

To straightforwardly view a coref document, copy & paste the tsv file to Spannotator. If you want to visualize the predictions from SpanBert, go to utils and run pred2tsv.py to generate the tsv file from predicted output json file.

Testing SpanBert

Go to utils and run python gum2conll.py to build up the dataset. It will generate train, dev (including by-genre set), and test set under ./dataset
Follow the instructions in SpanBert. Note that change the data directory.

Testing dcoref

Go to utils and run ./dcoref.sh

Results

Model	OntoNotes	OntoGUM
dcoref	57.8	39.7
SpanBert	79.6	64.6

Citations

@inproceedings{zhu-etal-2021-ontogum,
    title = "{O}nto{GUM}: Evaluating Contextualized {SOTA} Coreference Resolution on 12 More Genres",
    author = "Zhu, Yilun  and
      Pradhan, Sameer  and
      Zeldes, Amir",
    editor = "Zong, Chengqing  and
      Xia, Fei  and
      Li, Wenjie  and
      Navigli, Roberto",
    booktitle = "Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 2: Short Papers)",
    month = aug,
    year = "2021",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2021.acl-short.59",
    doi = "10.18653/v1/2021.acl-short.59",
    pages = "461--467",
}

@inproceedings{zhu-etal-2021-anatomy,
    title = "Anatomy of {O}nto{GUM}{---}{A}dapting {GUM} to the {O}nto{N}otes Scheme to Evaluate Robustness of {SOTA} Coreference Algorithms",
    author = "Zhu, Yilun  and
      Pradhan, Sameer  and
      Zeldes, Amir",
    editor = "Ogrodniczuk, Maciej  and
      Pradhan, Sameer  and
      Poesio, Massimo  and
      Grishina, Yulia  and
      Ng, Vincent",
    booktitle = "Proceedings of the Fourth Workshop on Computational Models of Reference, Anaphora and Coreference",
    month = nov,
    year = "2021",
    address = "Punta Cana, Dominican Republic",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2021.crac-1.15",
    doi = "10.18653/v1/2021.crac-1.15",
    pages = "141--149",
}

yilunzhu/ontogum