metadata

license: apache-2.0
task_categories:
  - feature-extraction
  - sentence-similarity
language:
  - ar
size_categories:
  - 10K<n<100K

Arabic Triplet with Multi Negatives

Dataset Summary

This dataset is a modified version of the Arabic subset of the Mr. TyDi dataset, tailored for retrieval and re-ranking tasks. The original dataset has been restructured by splitting the negative passages into separate fields (negative1, negative2, ..., negativeN) for each query. This modification allows more flexibility for training and evaluating retrieval and re-ranking models.

The dataset retains the original intent of Mr. Tydi, focusing on monolingual retrieval for the Arabic language while offering a new structure for ease of use in ranking-based learning tasks.

Dataset Structure

The dataset includes train split only where each query is paired with a set of positive passages and multiple individually enumerated negative passages (up to 30).

Example Data

Train Set

{
  "query_id": "1", 
  "query": "متى تم تطوير نظرية الحقل الكمي؟", 
  "positive_passages": [
    {
      "text": "بدأت نظرية الحقل الكمي بشكل طبيعي بدراسة التفاعلات الكهرومغناطيسية ..."
    }
  ],
  "negative1": {
      "text": "تم تنفيذ النهج مؤخرًا ليشمل نسخة جبرية من الحقل الكمي ..."
  },
  "negative2": {
      "text": "تتناول هذه المقالة الخلفية التاريخية لتطوير نظرية الحقل ..."
  },
  ...
}

Language Coverage

The dataset focuses exclusively on the Arabic subset of Mr. TyDi.

Loading the Dataset

You can load the dataset using the datasets library from Hugging Face:

from datasets import load_dataset

dataset = load_dataset('NAMAA-Space/Arabic-Triplet-With-Multi-Negatives')
dataset

Dataset Usage

The new format facilitates training retrieval and re-ranking models by providing explicit negative passage fields. This structure simplifies the handling of negative examples during model training and evaluation.

Citation Information

If you use this dataset in your research, please cite the original Mr. TyDi paper and this dataset as follows:

@article{mrtydi,
      title={{Mr. TyDi}: A Multi-lingual Benchmark for Dense Retrieval}, 
      author={Xinyu Zhang and Xueguang Ma and Peng Shi and Jimmy Lin},
      year={2021},
      journal={arXiv:2108.08787},
}

@dataset{Namaa,
      title={Arabic Triplet With Multi Negatives},
      author={Omer Nacar},
      year={2024},
      note={Hugging Face Dataset Repository}
}