The dataset could not be loaded because the splits use different data file formats, which is not supported. Read more about the splits configuration. Click for more details.

Couldn't infer the same data file format for all splits. Got {NamedSplit('train'): ('csv', {}), NamedSplit('validation'): (None, {}), NamedSplit('test'): ('csv', {})}

Error code:   FileFormatMismatchBetweenSplitsError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

Arabic End-of-Utterance Detection Dataset

Dataset Description

This dataset is designed for training Arabic End-of-Utterance (EOU) detection models for real-time voice agents. It contains complete and incomplete Arabic utterances derived from the SADA22 dataset with emphasis on Saudi dialect.

Dataset Structure

Files

train.csv: 125,155 samples (62,578 complete, 62,577 incomplete)
test.csv: 31,289 samples (15,644 complete, 15,645 incomplete)

Format

Each CSV contains:

text: Arabic utterance (3-50 words)
label: 0 (incomplete) or 1 (complete/EOU)

Sample

text	label
أريد أن أحجز موعد في العيادة غدا الساعة الثالثة	1
أريد أن أحجز موعد في العيادة	0

Dataset Creation

Source: SADA22 (Saudi Annotated Dataset for Arabic)

Methodology:

Extracted valid Arabic utterances (3-50 words)
Generated incomplete versions through random truncation (30-80%)
Balanced 1:1 ratio (complete:incomplete)
80/20 train/test split (stratified)

Usage

from datasets import load_dataset

training_set = hf_hub_download(
    repo_id="azeddinShr/arabic-eou-sada22",
    filename="train.csv",
    repo_type="dataset"
)
testing_set = hf_hub_download(
    repo_id="azeddinShr/arabic-eou-sada22",
    filename="test.csv",
    repo_type="dataset"
)

Citation

If you use this dataset, please cite the original SADA22 dataset:

@dataset{sada22,
  title={SADA22: Saudi Annotated Dataset for Arabic},
  author={Rashad, Mohamed},
  year={2022}
}

License

Same as source dataset (SADA22)

Downloads last month: 33