README.md

metadata

license: unknown
language:
  - en
task_categories:
  - text-classification

WI + Locness Task from the PEER Benchmark (Performance Evaluation of Edit Representations)

Description from the benchmark paper:

A dataset compiled by Bryant et al. (2019), built on top of: 1. A subset of the LOCNESS corpus (Granger 2014), which consists of essays written by native English students manually annotated with grammar errors. 2. Manually annotated examples from the Write & Improve online platform (Yannakoudakis et al. 2018). This dataset contains 3,600 annotated examples across three different CEFR levels (Little 2006): A (beginner), B (intermediate), and C (advanced). As with other datasets, we ignore examples where there are no grammar corrections. The data was originally published at https://doi.org/10.5281/zenodo.4478266.

Citations

PEER Benchmark:

@article{marrese-taylor-et-al-2021,
    title = {Variational Inference for Learning Representations of Natural Language Edits},
    volume = {35},
    url = {https://ojs.aaai.org/index.php/AAAI/article/view/17598}, DOI = {10.1609/aaai.v35i15.17598},
    number = {15},
    journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
    author = {Marrese-Taylor, Edison and Reid, Machel and Matsuo, Yutaka},
    year = {2021},
    month = {May},
    pages = {13552-13560},
}

Original data source:

@inproceedings{bryant-etal-2019-bea,
    title = "The {BEA}-2019 Shared Task on Grammatical Error Correction",
    author = "Bryant, Christopher  and
      Felice, Mariano  and
      Andersen, {\O}istein E.  and
      Briscoe, Ted",
    editor = "Yannakoudakis, Helen  and
      Kochmar, Ekaterina  and
      Leacock, Claudia  and
      Madnani, Nitin  and
      Pil{\'a}n, Ildik{\'o}  and
      Zesch, Torsten",
    booktitle = "Proceedings of the Fourteenth Workshop on Innovative Use of NLP for Building Educational Applications",
    month = aug,
    year = "2019",
    address = "Florence, Italy",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/W19-4406/",
    doi = "10.18653/v1/W19-4406",
    pages = "52--75",
    abstract = "This paper reports on the BEA-2019 Shared Task on Grammatical Error Correction (GEC). As with the CoNLL-2014 shared task, participants are required to correct all types of errors in test data. One of the main contributions of the BEA-2019 shared task is the introduction of a new dataset, the Write{\&}Improve+LOCNESS corpus, which represents a wider range of native and learner English levels and abilities. Another contribution is the introduction of tracks, which control the amount of annotated data available to participants. Systems are evaluated in terms of ERRANT F{\_}0.5, which allows us to report a much wider range of performance statistics. The competition was hosted on Codalab and remains open for further submissions on the blind test set."
}