README.md

metadata

dataset_info:
  features:
    - name: pair
      dtype: string
    - name: label
      dtype: int64
  splits:
    - name: train
      num_bytes: 18442362
      num_examples: 16205
    - name: test
      num_bytes: 4165369
      num_examples: 3717
    - name: val
      num_bytes: 531051
      num_examples: 778
  download_size: 2063012
  dataset_size: 23138782
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
      - split: val
        path: data/val-*
license: apache-2.0
task_categories:
  - text-classification
  - token-classification
language:
  - ar

AraEventCoref

AraEventCoref is an Arabic event coreference dataset comprising 50 annotated news articles with 12,069 tokens, 1,381 events, and 159 coreference chains. Target events are surrounded by position-aware markers <<حدث>> before and after each event. The label 1 indicates a coreference, while 0 indicates no coreference. The dataset features high annotation reliability with a CoNLL score of 75.8% and an inter-annotator agreement of 96% for event triggers.

• Curated by: Dr. Mohammed Aldawsari
• Language(s) (NLP): Arabic

Source Data

SANAD dataset

Dataset Card Contact

mohammed.aldawsari@psau.edu.sa