---
language:
- en
license:
- cc-by-sa-4.0
multilinguality:
- monolingual
paperswithcode_id: beir
pretty_name: BEIR Benchmark
task_categories:
- zero-shot-classification
- text-retrieval
task_ids:
- document-retrieval
- entity-linking-retrieval
- fact-checking-retrieval
tags:
- biomedical-information-retrieval
- citation-prediction-retrieval
- passage-retrieval
- news-retrieval
- argument-retrieval
- zero-shot-information-retrieval
- tweet-retrieval
- question-answering-retrieval
- duplication-question-retrieval
- zero-shot-retrieval
configs:
- config_name: android
  data_files:
  - split: corpus
    path: android/corpus/corpus-*
  - split: queries
    path: android/queries/queries-*
- config_name: english
  data_files:
  - split: corpus
    path: english/corpus/corpus-*
  - split: queries
    path: english/queries/queries-*
- config_name: gaming
  data_files:
  - split: corpus
    path: gaming/corpus/corpus-*
  - split: queries
    path: gaming/queries/queries-*
- config_name: gis
  data_files:
  - split: corpus
    path: gis/corpus/corpus-*
  - split: queries
    path: gis/queries/queries-*
- config_name: mathematica
  data_files:
  - split: corpus
    path: mathematica/corpus/corpus-*
  - split: queries
    path: mathematica/queries/queries-*
- config_name: physics
  data_files:
  - split: corpus
    path: physics/corpus/corpus-*
  - split: queries
    path: physics/queries/queries-*
- config_name: programmers
  data_files:
  - split: corpus
    path: programmers/corpus/corpus-*
  - split: queries
    path: programmers/queries/queries-*
- config_name: stats
  data_files:
  - split: corpus
    path: stats/corpus/corpus-*
  - split: queries
    path: stats/queries/queries-*
- config_name: tex
  data_files:
  - split: corpus
    path: tex/corpus/corpus-*
  - split: queries
    path: tex/queries/queries-*
- config_name: unix
  data_files:
  - split: corpus
    path: unix/corpus/corpus-*
  - split: queries
    path: unix/queries/queries-*
- config_name: webmasters
  data_files:
  - split: corpus
    path: webmasters/corpus/corpus-*
  - split: queries
    path: webmasters/queries/queries-*
- config_name: wordpress
  data_files:
  - split: corpus
    path: wordpress/corpus/corpus-*
  - split: queries
    path: wordpress/queries/queries-*
dataset_info:
- config_name: android
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: corpus
    num_bytes: 8255816
    num_examples: 22998
  - name: queries
    num_bytes: 31997
    num_examples: 699
  download_size: 8287813
  dataset_size: 8287813
- config_name: english
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: corpus
    num_bytes: 12417951
    num_examples: 40221
  - name: queries
    num_bytes: 68095
    num_examples: 1570
  download_size: 12486046
  dataset_size: 12486046
- config_name: gaming
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: corpus
    num_bytes: 14156689
    num_examples: 45301
  - name: queries
    num_bytes: 70040
    num_examples: 1595
  download_size: 14226729
  dataset_size: 14226729
- config_name: gis
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: corpus
    num_bytes: 21024580
    num_examples: 37637
  - name: queries
    num_bytes: 39653
    num_examples: 885
  download_size: 21064233
  dataset_size: 21064233
- config_name: mathematica
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: corpus
    num_bytes: 10765498
    num_examples: 16705
  - name: queries
    num_bytes: 34583
    num_examples: 804
  download_size: 10800081
  dataset_size: 10800081
- config_name: physics
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: corpus
    num_bytes: 18811650
    num_examples: 38316
  - name: queries
    num_bytes: 47465
    num_examples: 1039
  download_size: 18859115
  dataset_size: 18859115
- config_name: programmers
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: corpus
    num_bytes: 20111039
    num_examples: 32176
  - name: queries
    num_bytes: 41633
    num_examples: 876
  download_size: 20152672
  dataset_size: 20152672
- config_name: stats
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: corpus
    num_bytes: 25668372
    num_examples: 42269
  - name: queries
    num_bytes: 30347
    num_examples: 652
  download_size: 25698719
  dataset_size: 25698719
- config_name: tex
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: corpus
    num_bytes: 44756206
    num_examples: 68184
  - name: queries
    num_bytes: 117655
    num_examples: 2906
  download_size: 44873861
  dataset_size: 44873861
- config_name: unix
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: corpus
    num_bytes: 25746785
    num_examples: 47382
  - name: queries
    num_bytes: 47358
    num_examples: 1072
  download_size: 25794143
  dataset_size: 25794143
- config_name: webmasters
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: corpus
    num_bytes: 7353229
    num_examples: 17405
  - name: queries
    num_bytes: 24033
    num_examples: 506
  download_size: 7377262
  dataset_size: 7377262
- config_name: wordpress
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: corpus
    num_bytes: 27597725
    num_examples: 48605
  - name: queries
    num_bytes: 24475
    num_examples: 541
  download_size: 27622200
  dataset_size: 27622200
---

# Dataset Card for BEIR Benchmark

## Dataset Description

- **Homepage:** https://beir.ai
- **Repository:** https://beir.ai
- **Paper:** https://openreview.net/forum?id=wCu6T5xFjeJ
- **Leaderboard:** https://docs.google.com/spreadsheets/d/1L8aACyPaXrL8iEelJLGqlMqXKPX2oSP_R10pZoy77Ns
- **Point of Contact:** nandan.thakur@uwaterloo.ca

### Dataset Summary

BEIR is a heterogeneous benchmark built from 18 diverse datasets representing 9 information retrieval tasks.

- Fact-checking: [FEVER](http://fever.ai), [Climate-FEVER](http://climatefever.ai), [SciFact](https://github.com/allenai/scifact)
- Question-Answering: [NQ](https://ai.google.com/research/NaturalQuestions), [HotpotQA](https://hotpotqa.github.io), [FiQA-2018](https://sites.google.com/view/fiqa/)
- Bio-Medical IR: [TREC-COVID](https://ir.nist.gov/covidSubmit/index.html), [BioASQ](http://bioasq.org), [NFCorpus](https://www.cl.uni-heidelberg.de/statnlpgroup/nfcorpus/)
- News Retrieval: [TREC-NEWS](https://trec.nist.gov/data/news2019.html), [Robust04](https://trec.nist.gov/data/robust/04.guidelines.html)
- Argument Retrieval: [Touche-2020](https://webis.de/events/touche-20/shared-task-1.html), [ArguAna](tp://argumentation.bplaced.net/arguana/data)
- Duplicate Question Retrieval: [Quora](https://www.quora.com/q/quoradata/First-Quora-Dataset-Release-Question-Pairs), [CqaDupstack](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/)
- Citation-Prediction: [SCIDOCS](https://allenai.org/data/scidocs)
- Tweet Retrieval: [Signal-1M](https://research.signal-ai.com/datasets/signal1m-tweetir.html)
- Entity Retrieval: [DBPedia](https://github.com/iai-group/DBpedia-Entity/)

> **This `cqadupstack` subset is the Duplicate Question Retrieval task within BEIR, with 12 domains.**

### Languages

All tasks are in English (`en`).

## Dataset Structure

This dataset uses the standard BEIR retrieval layout with 12 domain configs:
`android`, `english`, `gaming`, `gis`, `mathematica`, `physics`, `programmers`, `stats`, `tex`, `unix`, `webmasters`, `wordpress`.

Each config includes:

- `corpus`: one row per document with `_id`, `title`, `text`
- `queries`: one row per query with `_id`, `title`, `text`

### Data Fields

- `_id` (`string`): unique identifier
- `title` (`string`): title (empty string when unavailable)
- `text` (`string`): document/query text

### CQADupstack Data Splits

| Domain | Split | Rows |
| --- | --- | ---: |
| `android` | corpus | 22,998 |
| `android` | queries | 699 |
| `english` | corpus | 40,221 |
| `english` | queries | 1,570 |
| `gaming` | corpus | 45,301 |
| `gaming` | queries | 1,595 |
| `gis` | corpus | 37,637 |
| `gis` | queries | 885 |
| `mathematica` | corpus | 16,705 |
| `mathematica` | queries | 804 |
| `physics` | corpus | 38,316 |
| `physics` | queries | 1,039 |
| `programmers` | corpus | 32,176 |
| `programmers` | queries | 876 |
| `stats` | corpus | 42,269 |
| `stats` | queries | 652 |
| `tex` | corpus | 68,184 |
| `tex` | queries | 2,906 |
| `unix` | corpus | 47,382 |
| `unix` | queries | 1,072 |
| `webmasters` | corpus | 17,405 |
| `webmasters` | queries | 506 |
| `wordpress` | corpus | 48,605 |
| `wordpress` | queries | 541 |

## Citation Information

```bibtex
@inproceedings{
thakur2021beir,
title={{BEIR}: A Heterogeneous Benchmark for Zero-shot Evaluation of Information Retrieval Models},
author={Nandan Thakur and Nils Reimers and Andreas R{\"u}ckl{\'e} and Abhishek Srivastava and Iryna Gurevych},
booktitle={Thirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 2)},
year={2021},
url={https://openreview.net/forum?id=wCu6T5xFjeJ}
}
```