license: cc-by-nc-sa-4.0
task_categories:
- text-classification
language:
- en
pretty_name: KUPA-KEYS
size_categories:
- 1K<n<10K
This repository hosts the dataset collected during the project, 'Deep Learning for Language Assessment', as detailed in the paper "Logging Keystrokes in Writing by English Learners", to appear in the proceedings of LREC-COLING 2024.
The dataset is named KUPA-KEYS (King's College London & Université Paris Cité Keys). It contains texts written by 1,006 participants in our crowdsourcing study, recruited on Prolific. Task 1 involved a text-copy task; Task 2 involved essay writing in response to a 'Just for Fun' prompt from Write & Improve (W&I), used with permission. Keystroke data for these texts are included in the dataset, as well as metadata and CEFR level grades for the free-text essays. Further details about the data collection process, annotation and analysis may be found in our LREC-COLING paper.
Contents:
- KUPA-KEYS-META.csv : information about each participant, including computing environment & keyboard layout, education & level of English, task 1 and task 2 statistics, the essay prompt for task 2, the final form of their task 2 essay, and the scores / CEFR levels received from human markers (h1, h2, h3) and the W&I automarker (a0).
- KUPA-KEYS-TASK-1.csv : all keystroke events for each participant undertaking task 1, the text-copy task.
- KUPA-KEYS-TASK-2.csv : all keystroke events for each participant undertaking task 2, the essay writing task.
For more information about the contents of the files, see our paper forthcoming at LREC-COLING 2024, or the DatasetDescription page.
Georgios Velentzas, Andrew Caines, Rita Borgo, Erin Pacquetet, Clive Hamilton, Taylor Arnold, Diane Nicholls, Paula Buttery, Thomas Gaillat, Nicolas Ballier and Helen Yannakoudakis