---
dataset_info:
- config_name: corpus
  features:
  - name: _id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 6809516
    num_examples: 2919
  download_size: 2948742
  dataset_size: 6809516
- config_name: qrels
  features:
  - name: query-id
    dtype: string
  - name: corpus-id
    dtype: string
  splits:
  - name: train
    num_bytes: 1054
    num_examples: 56
  download_size: 2081
  dataset_size: 1054
- config_name: queries
  features:
  - name: _id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 8645
    num_examples: 50
  download_size: 6492
  dataset_size: 8645
configs:
- config_name: corpus
  data_files:
  - split: train
    path: corpus/train-*
- config_name: qrels
  data_files:
  - split: train
    path: qrels/train-*
- config_name: queries
  data_files:
  - split: train
    path: queries/train-*
language:
- ar
tags:
- information-retrieval
- nanobeir
- benchmark
---

# NanoSciFact - Arabic

This dataset is the Arabic version of the NanoSciFact benchmark from the NanoBEIR multilingual collection.

## Dataset Origin

This dataset is derived from [lightonai/nanobeir-multilingual](https://huggingface.co/datasets/lightonai/nanobeir-multilingual).

NanoBEIR is a smaller version of the BEIR benchmark designed for efficient evaluation of information retrieval models.

## Dataset Structure

The dataset contains three configurations:

- **corpus**: The document collection to search through
- **queries**: The search queries
- **qrels**: Relevance judgments (query-document pairs with relevance scores)

## Usage

```python
from datasets import load_dataset

# Load the different configurations
corpus = load_dataset("wissamantoun/NanoSciFact_Arabic", "corpus")
queries = load_dataset("wissamantoun/NanoSciFact_Arabic", "queries")
qrels = load_dataset("wissamantoun/NanoSciFact_Arabic", "qrels")
```

## Citation

If you use this dataset, please cite the original BEIR and NanoBEIR work.