metadata

language:
  - ur
license: cc-by-4.0
multilinguality:
  - monolingual
source_datasets:
  - NanoArguAna
task_categories:
  - text-retrieval
task_ids:
  - document-retrieval
tags:
  - text-retrieval
dataset_info:
  - config_name: corpus
    features:
      - name: _id
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 6404731
        num_examples: 3635
    download_size: 2885131
    dataset_size: 6404731
  - config_name: qrels
    features:
      - name: query-id
        dtype: string
      - name: corpus-id
        dtype: string
    splits:
      - name: train
        num_bytes: 3496
        num_examples: 50
    download_size: 3187
    dataset_size: 3496
  - config_name: queries
    features:
      - name: _id
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 104181
        num_examples: 50
    download_size: 57602
    dataset_size: 104181
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-*

Bharat-NanoBEIR: Indian Language Information Retrieval Dataset

Overview

This dataset is part of the Bharat-NanoBEIR collection, which provides information retrieval datasets for Indian languages. It is derived from the NanoBEIR project, which offers smaller versions of BEIR datasets containing 50 queries and up to 10K documents each.

Dataset Description

This particular dataset is the Urdu version of the NanoArguAna dataset, specifically adapted for information retrieval tasks. The translation and adaptation maintain the core structure of the original NanoBEIR while making it accessible for Urdu language processing.

Usage

This dataset is designed for:

Information Retrieval (IR) system development in Urdu
Evaluation of multilingual search capabilities
Cross-lingual information retrieval research
Benchmarking Urdu language models for search tasks

Dataset Structure

The dataset consists of three main components:

Corpus: Collection of documents in Urdu
Queries: Search queries in Urdu
QRels: Relevance judgments connecting queries to relevant documents

Citation

If you use this dataset, please cite:

@misc{bharat-nanobeir,
  title={Bharat-NanoBEIR: Indian Language Information Retrieval Datasets},
  year={2024},
  url={https://huggingface.co/datasets/carlfeynman/Bharat_NanoArguAna_ur}
}

Additional Information

Language: Urdu (ur)
License: CC-BY-4.0
Original Dataset: NanoBEIR
Domain: Information Retrieval

License

This dataset is licensed under CC-BY-4.0. Please see the LICENSE file for details.