nthakur commited on
Commit
6cd1d4c
1 Parent(s): c376963

initial file add and README.

Browse files
Files changed (2) hide show
  1. README.md +510 -0
  2. train.jsonl.gz +3 -0
README.md ADDED
@@ -0,0 +1,510 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ annotations_creators: []
3
+ language_creators: []
4
+ language:
5
+ - en
6
+ license:
7
+ - cc-by-sa-4.0
8
+ multilinguality:
9
+ - monolingual
10
+ paperswithcode_id: beir
11
+ pretty_name: BEIR Benchmark
12
+ size_categories:
13
+ msmarco:
14
+ - 1M<n<10M
15
+ trec-covid:
16
+ - 100k<n<1M
17
+ nfcorpus:
18
+ - 1K<n<10K
19
+ nq:
20
+ - 1M<n<10M
21
+ hotpotqa:
22
+ - 1M<n<10M
23
+ fiqa:
24
+ - 10K<n<100K
25
+ arguana:
26
+ - 1K<n<10K
27
+ touche-2020:
28
+ - 100K<n<1M
29
+ cqadupstack:
30
+ - 100K<n<1M
31
+ quora:
32
+ - 100K<n<1M
33
+ dbpedia:
34
+ - 1M<n<10M
35
+ scidocs:
36
+ - 10K<n<100K
37
+ fever:
38
+ - 1M<n<10M
39
+ climate-fever:
40
+ - 1M<n<10M
41
+ scifact:
42
+ - 1K<n<10K
43
+ source_datasets: []
44
+ task_categories:
45
+ - text-retrieval
46
+ ---
47
+
48
+ # NFCorpus: 20 generated queries (BEIR Benchmark)
49
+
50
+ This HF dataset contains the top-20 synthetic queries generated for each passage in the above BEIR benchmark dataset.
51
+
52
+ - DocT5query model used: [BeIR/query-gen-msmarco-t5-base-v1](https://huggingface.co/BeIR/query-gen-msmarco-t5-base-v1)
53
+ - id (str): unique document id in NFCorpus in the BEIR benchmark (`corpus.jsonl`).
54
+ - Questions generated: 20
55
+ - Code used for generation: [evaluate_anserini_docT5query_parallel.py](https://github.com/beir-cellar/beir/blob/main/examples/retrieval/evaluation/sparse/evaluate_anserini_docT5query_parallel.py)
56
+
57
+
58
+ Below contains the old dataset card for the BEIR benchmark.
59
+
60
+
61
+ # Dataset Card for BEIR Benchmark
62
+
63
+ ## Table of Contents
64
+ - [Dataset Description](#dataset-description)
65
+ - [Dataset Summary](#dataset-summary)
66
+ - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
67
+ - [Languages](#languages)
68
+ - [Dataset Structure](#dataset-structure)
69
+ - [Data Instances](#data-instances)
70
+ - [Data Fields](#data-fields)
71
+ - [Data Splits](#data-splits)
72
+ - [Dataset Creation](#dataset-creation)
73
+ - [Curation Rationale](#curation-rationale)
74
+ - [Source Data](#source-data)
75
+ - [Annotations](#annotations)
76
+ - [Personal and Sensitive Information](#personal-and-sensitive-information)
77
+ - [Considerations for Using the Data](#considerations-for-using-the-data)
78
+ - [Social Impact of Dataset](#social-impact-of-dataset)
79
+ - [Discussion of Biases](#discussion-of-biases)
80
+ - [Other Known Limitations](#other-known-limitations)
81
+ - [Additional Information](#additional-information)
82
+ - [Dataset Curators](#dataset-curators)
83
+ - [Licensing Information](#licensing-information)
84
+ - [Citation Information](#citation-information)
85
+ - [Contributions](#contributions)
86
+
87
+ ## Dataset Description
88
+
89
+ - **Homepage:** https://github.com/UKPLab/beir
90
+ - **Repository:** https://github.com/UKPLab/beir
91
+ - **Paper:** https://openreview.net/forum?id=wCu6T5xFjeJ
92
+ - **Leaderboard:** https://docs.google.com/spreadsheets/d/1L8aACyPaXrL8iEelJLGqlMqXKPX2oSP_R10pZoy77Ns
93
+ - **Point of Contact:** nandan.thakur@uwaterloo.ca
94
+
95
+ ### Dataset Summary
96
+
97
+ BEIR is a heterogeneous benchmark that has been built from 18 diverse datasets representing 9 information retrieval tasks:
98
+
99
+ - Fact-checking: [FEVER](http://fever.ai), [Climate-FEVER](http://climatefever.ai), [SciFact](https://github.com/allenai/scifact)
100
+ - Question-Answering: [NQ](https://ai.google.com/research/NaturalQuestions), [HotpotQA](https://hotpotqa.github.io), [FiQA-2018](https://sites.google.com/view/fiqa/)
101
+ - 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/)
102
+ - News Retrieval: [TREC-NEWS](https://trec.nist.gov/data/news2019.html), [Robust04](https://trec.nist.gov/data/robust/04.guidelines.html)
103
+ - Argument Retrieval: [Touche-2020](https://webis.de/events/touche-20/shared-task-1.html), [ArguAna](tp://argumentation.bplaced.net/arguana/data)
104
+ - 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/)
105
+ - Citation-Prediction: [SCIDOCS](https://allenai.org/data/scidocs)
106
+ - Tweet Retrieval: [Signal-1M](https://research.signal-ai.com/datasets/signal1m-tweetir.html)
107
+ - Entity Retrieval: [DBPedia](https://github.com/iai-group/DBpedia-Entity/)
108
+
109
+ All these datasets have been preprocessed and can be used for your experiments.
110
+
111
+
112
+ ```python
113
+
114
+ ```
115
+
116
+ ### Supported Tasks and Leaderboards
117
+
118
+ The dataset supports a leaderboard that evaluates models against task-specific metrics such as F1 or EM, as well as their ability to retrieve supporting information from Wikipedia.
119
+
120
+ The current best performing models can be found [here](https://eval.ai/web/challenges/challenge-page/689/leaderboard/).
121
+
122
+ ### Languages
123
+
124
+ All tasks are in English (`en`).
125
+
126
+ ## Dataset Structure
127
+
128
+ All BEIR datasets must contain a corpus, queries and qrels (relevance judgments file). They must be in the following format:
129
+ - `corpus` file: a `.jsonl` file (jsonlines) that contains a list of dictionaries, each with three fields `_id` with unique document identifier, `title` with document title (optional) and `text` with document paragraph or passage. For example: `{"_id": "doc1", "title": "Albert Einstein", "text": "Albert Einstein was a German-born...."}`
130
+ - `queries` file: a `.jsonl` file (jsonlines) that contains a list of dictionaries, each with two fields `_id` with unique query identifier and `text` with query text. For example: `{"_id": "q1", "text": "Who developed the mass-energy equivalence formula?"}`
131
+ - `qrels` file: a `.tsv` file (tab-seperated) that contains three columns, i.e. the `query-id`, `corpus-id` and `score` in this order. Keep 1st row as header. For example: `q1 doc1 1`
132
+
133
+ ### Data Instances
134
+
135
+ A high level example of any beir dataset:
136
+
137
+ ```python
138
+ corpus = {
139
+ "doc1" : {
140
+ "title": "Albert Einstein",
141
+ "text": "Albert Einstein was a German-born theoretical physicist. who developed the theory of relativity, \
142
+ one of the two pillars of modern physics (alongside quantum mechanics). His work is also known for \
143
+ its influence on the philosophy of science. He is best known to the general public for his mass–energy \
144
+ equivalence formula E = mc2, which has been dubbed 'the world's most famous equation'. He received the 1921 \
145
+ Nobel Prize in Physics 'for his services to theoretical physics, and especially for his discovery of the law \
146
+ of the photoelectric effect', a pivotal step in the development of quantum theory."
147
+ },
148
+ "doc2" : {
149
+ "title": "", # Keep title an empty string if not present
150
+ "text": "Wheat beer is a top-fermented beer which is brewed with a large proportion of wheat relative to the amount of \
151
+ malted barley. The two main varieties are German Weißbier and Belgian witbier; other types include Lambic (made\
152
+ with wild yeast), Berliner Weisse (a cloudy, sour beer), and Gose (a sour, salty beer)."
153
+ },
154
+ }
155
+
156
+ queries = {
157
+ "q1" : "Who developed the mass-energy equivalence formula?",
158
+ "q2" : "Which beer is brewed with a large proportion of wheat?"
159
+ }
160
+
161
+ qrels = {
162
+ "q1" : {"doc1": 1},
163
+ "q2" : {"doc2": 1},
164
+ }
165
+ ```
166
+
167
+ ### Data Fields
168
+
169
+ Examples from all configurations have the following features:
170
+
171
+ ### Corpus
172
+ - `corpus`: a `dict` feature representing the document title and passage text, made up of:
173
+ - `_id`: a `string` feature representing the unique document id
174
+ - `title`: a `string` feature, denoting the title of the document.
175
+ - `text`: a `string` feature, denoting the text of the document.
176
+
177
+ ### Queries
178
+ - `queries`: a `dict` feature representing the query, made up of:
179
+ - `_id`: a `string` feature representing the unique query id
180
+ - `text`: a `string` feature, denoting the text of the query.
181
+
182
+ ### Qrels
183
+ - `qrels`: a `dict` feature representing the query document relevance judgements, made up of:
184
+ - `_id`: a `string` feature representing the query id
185
+ - `_id`: a `string` feature, denoting the document id.
186
+ - `score`: a `int32` feature, denoting the relevance judgement between query and document.
187
+
188
+
189
+ ### Data Splits
190
+
191
+ | Dataset | Website| BEIR-Name | Type | Queries | Corpus | Rel D/Q | Down-load | md5 |
192
+ | -------- | -----| ---------| --------- | ----------- | ---------| ---------| :----------: | :------:|
193
+ | MSMARCO | [Homepage](https://microsoft.github.io/msmarco/)| ``msmarco`` | ``train``<br>``dev``<br>``test``| 6,980 | 8.84M | 1.1 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/msmarco.zip) | ``444067daf65d982533ea17ebd59501e4`` |
194
+ | TREC-COVID | [Homepage](https://ir.nist.gov/covidSubmit/index.html)| ``trec-covid``| ``test``| 50| 171K| 493.5 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/trec-covid.zip) | ``ce62140cb23feb9becf6270d0d1fe6d1`` |
195
+ | NFCorpus | [Homepage](https://www.cl.uni-heidelberg.de/statnlpgroup/nfcorpus/) | ``nfcorpus`` | ``train``<br>``dev``<br>``test``| 323 | 3.6K | 38.2 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/nfcorpus.zip) | ``a89dba18a62ef92f7d323ec890a0d38d`` |
196
+ | BioASQ | [Homepage](http://bioasq.org) | ``bioasq``| ``train``<br>``test`` | 500 | 14.91M | 8.05 | No | [How to Reproduce?](https://github.com/UKPLab/beir/blob/main/examples/dataset#2-bioasq) |
197
+ | NQ | [Homepage](https://ai.google.com/research/NaturalQuestions) | ``nq``| ``train``<br>``test``| 3,452 | 2.68M | 1.2 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/nq.zip) | ``d4d3d2e48787a744b6f6e691ff534307`` |
198
+ | HotpotQA | [Homepage](https://hotpotqa.github.io) | ``hotpotqa``| ``train``<br>``dev``<br>``test``| 7,405 | 5.23M | 2.0 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/hotpotqa.zip) | ``f412724f78b0d91183a0e86805e16114`` |
199
+ | FiQA-2018 | [Homepage](https://sites.google.com/view/fiqa/) | ``fiqa`` | ``train``<br>``dev``<br>``test``| 648 | 57K | 2.6 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/fiqa.zip) | ``17918ed23cd04fb15047f73e6c3bd9d9`` |
200
+ | Signal-1M(RT) | [Homepage](https://research.signal-ai.com/datasets/signal1m-tweetir.html)| ``signal1m`` | ``test``| 97 | 2.86M | 19.6 | No | [How to Reproduce?](https://github.com/UKPLab/beir/blob/main/examples/dataset#4-signal-1m) |
201
+ | TREC-NEWS | [Homepage](https://trec.nist.gov/data/news2019.html) | ``trec-news`` | ``test``| 57 | 595K | 19.6 | No | [How to Reproduce?](https://github.com/UKPLab/beir/blob/main/examples/dataset#1-trec-news) |
202
+ | ArguAna | [Homepage](http://argumentation.bplaced.net/arguana/data) | ``arguana``| ``test`` | 1,406 | 8.67K | 1.0 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/arguana.zip) | ``8ad3e3c2a5867cdced806d6503f29b99`` |
203
+ | Touche-2020| [Homepage](https://webis.de/events/touche-20/shared-task-1.html) | ``webis-touche2020``| ``test``| 49 | 382K | 19.0 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/webis-touche2020.zip) | ``46f650ba5a527fc69e0a6521c5a23563`` |
204
+ | CQADupstack| [Homepage](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | ``cqadupstack``| ``test``| 13,145 | 457K | 1.4 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/cqadupstack.zip) | ``4e41456d7df8ee7760a7f866133bda78`` |
205
+ | Quora| [Homepage](https://www.quora.com/q/quoradata/First-Quora-Dataset-Release-Question-Pairs) | ``quora``| ``dev``<br>``test``| 10,000 | 523K | 1.6 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/quora.zip) | ``18fb154900ba42a600f84b839c173167`` |
206
+ | DBPedia | [Homepage](https://github.com/iai-group/DBpedia-Entity/) | ``dbpedia-entity``| ``dev``<br>``test``| 400 | 4.63M | 38.2 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/dbpedia-entity.zip) | ``c2a39eb420a3164af735795df012ac2c`` |
207
+ | SCIDOCS| [Homepage](https://allenai.org/data/scidocs) | ``scidocs``| ``test``| 1,000 | 25K | 4.9 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/scidocs.zip) | ``38121350fc3a4d2f48850f6aff52e4a9`` |
208
+ | FEVER | [Homepage](http://fever.ai) | ``fever``| ``train``<br>``dev``<br>``test``| 6,666 | 5.42M | 1.2| [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/fever.zip) | ``5a818580227bfb4b35bb6fa46d9b6c03`` |
209
+ | Climate-FEVER| [Homepage](http://climatefever.ai) | ``climate-fever``|``test``| 1,535 | 5.42M | 3.0 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/climate-fever.zip) | ``8b66f0a9126c521bae2bde127b4dc99d`` |
210
+ | SciFact| [Homepage](https://github.com/allenai/scifact) | ``scifact``| ``train``<br>``test``| 300 | 5K | 1.1 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/scifact.zip) | ``5f7d1de60b170fc8027bb7898e2efca1`` |
211
+ | Robust04 | [Homepage](https://trec.nist.gov/data/robust/04.guidelines.html) | ``robust04``| ``test``| 249 | 528K | 69.9 | No | [How to Reproduce?](https://github.com/UKPLab/beir/blob/main/examples/dataset#3-robust04) |
212
+
213
+
214
+ ## Dataset Creation
215
+
216
+ ### Curation Rationale
217
+
218
+ [Needs More Information]
219
+
220
+ ### Source Data
221
+
222
+ #### Initial Data Collection and Normalization
223
+
224
+ [Needs More Information]
225
+
226
+ #### Who are the source language producers?
227
+
228
+ [Needs More Information]
229
+
230
+ ### Annotations
231
+
232
+ #### Annotation process
233
+
234
+ [Needs More Information]
235
+
236
+ #### Who are the annotators?
237
+
238
+ [Needs More Information]
239
+
240
+ ### Personal and Sensitive Information
241
+
242
+ [Needs More Information]
243
+
244
+ ## Considerations for Using the Data
245
+
246
+ ### Social Impact of Dataset
247
+
248
+ [Needs More Information]
249
+
250
+ ### Discussion of Biases
251
+
252
+ [Needs More Information]
253
+
254
+ ### Other Known Limitations
255
+
256
+ [Needs More Information]
257
+
258
+ ## Additional Information
259
+
260
+ ### Dataset Curators
261
+
262
+ [Needs More Information]
263
+
264
+ ### Licensing Information
265
+
266
+ [Needs More Information]
267
+
268
+ ### Citation Information
269
+
270
+ Cite as:
271
+ ```
272
+ @inproceedings{
273
+ thakur2021beir,
274
+ title={{BEIR}: A Heterogeneous Benchmark for Zero-shot Evaluation of Information Retrieval Models},
275
+ author={Nandan Thakur and Nils Reimers and Andreas R{\"u}ckl{\'e} and Abhishek Srivastava and Iryna Gurevych},
276
+ booktitle={Thirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 2)},
277
+ year={2021},
278
+ url={https://openreview.net/forum?id=wCu6T5xFjeJ}
279
+ }
280
+ ```
281
+
282
+ ### Contributions
283
+
284
+ Thanks to [@Nthakur20](https://github.com/Nthakur20) for adding this dataset.Top-20 generated queries for every passage in NFCorpus
285
+
286
+
287
+ # Dataset Card for BEIR Benchmark
288
+
289
+ ## Table of Contents
290
+ - [Dataset Description](#dataset-description)
291
+ - [Dataset Summary](#dataset-summary)
292
+ - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
293
+ - [Languages](#languages)
294
+ - [Dataset Structure](#dataset-structure)
295
+ - [Data Instances](#data-instances)
296
+ - [Data Fields](#data-fields)
297
+ - [Data Splits](#data-splits)
298
+ - [Dataset Creation](#dataset-creation)
299
+ - [Curation Rationale](#curation-rationale)
300
+ - [Source Data](#source-data)
301
+ - [Annotations](#annotations)
302
+ - [Personal and Sensitive Information](#personal-and-sensitive-information)
303
+ - [Considerations for Using the Data](#considerations-for-using-the-data)
304
+ - [Social Impact of Dataset](#social-impact-of-dataset)
305
+ - [Discussion of Biases](#discussion-of-biases)
306
+ - [Other Known Limitations](#other-known-limitations)
307
+ - [Additional Information](#additional-information)
308
+ - [Dataset Curators](#dataset-curators)
309
+ - [Licensing Information](#licensing-information)
310
+ - [Citation Information](#citation-information)
311
+ - [Contributions](#contributions)
312
+
313
+ ## Dataset Description
314
+
315
+ - **Homepage:** https://github.com/UKPLab/beir
316
+ - **Repository:** https://github.com/UKPLab/beir
317
+ - **Paper:** https://openreview.net/forum?id=wCu6T5xFjeJ
318
+ - **Leaderboard:** https://docs.google.com/spreadsheets/d/1L8aACyPaXrL8iEelJLGqlMqXKPX2oSP_R10pZoy77Ns
319
+ - **Point of Contact:** nandan.thakur@uwaterloo.ca
320
+
321
+ ### Dataset Summary
322
+
323
+ BEIR is a heterogeneous benchmark that has been built from 18 diverse datasets representing 9 information retrieval tasks:
324
+
325
+ - Fact-checking: [FEVER](http://fever.ai), [Climate-FEVER](http://climatefever.ai), [SciFact](https://github.com/allenai/scifact)
326
+ - Question-Answering: [NQ](https://ai.google.com/research/NaturalQuestions), [HotpotQA](https://hotpotqa.github.io), [FiQA-2018](https://sites.google.com/view/fiqa/)
327
+ - 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/)
328
+ - News Retrieval: [TREC-NEWS](https://trec.nist.gov/data/news2019.html), [Robust04](https://trec.nist.gov/data/robust/04.guidelines.html)
329
+ - Argument Retrieval: [Touche-2020](https://webis.de/events/touche-20/shared-task-1.html), [ArguAna](tp://argumentation.bplaced.net/arguana/data)
330
+ - 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/)
331
+ - Citation-Prediction: [SCIDOCS](https://allenai.org/data/scidocs)
332
+ - Tweet Retrieval: [Signal-1M](https://research.signal-ai.com/datasets/signal1m-tweetir.html)
333
+ - Entity Retrieval: [DBPedia](https://github.com/iai-group/DBpedia-Entity/)
334
+
335
+ All these datasets have been preprocessed and can be used for your experiments.
336
+
337
+
338
+ ```python
339
+
340
+ ```
341
+
342
+ ### Supported Tasks and Leaderboards
343
+
344
+ The dataset supports a leaderboard that evaluates models against task-specific metrics such as F1 or EM, as well as their ability to retrieve supporting information from Wikipedia.
345
+
346
+ The current best performing models can be found [here](https://eval.ai/web/challenges/challenge-page/689/leaderboard/).
347
+
348
+ ### Languages
349
+
350
+ All tasks are in English (`en`).
351
+
352
+ ## Dataset Structure
353
+
354
+ All BEIR datasets must contain a corpus, queries and qrels (relevance judgments file). They must be in the following format:
355
+ - `corpus` file: a `.jsonl` file (jsonlines) that contains a list of dictionaries, each with three fields `_id` with unique document identifier, `title` with document title (optional) and `text` with document paragraph or passage. For example: `{"_id": "doc1", "title": "Albert Einstein", "text": "Albert Einstein was a German-born...."}`
356
+ - `queries` file: a `.jsonl` file (jsonlines) that contains a list of dictionaries, each with two fields `_id` with unique query identifier and `text` with query text. For example: `{"_id": "q1", "text": "Who developed the mass-energy equivalence formula?"}`
357
+ - `qrels` file: a `.tsv` file (tab-seperated) that contains three columns, i.e. the `query-id`, `corpus-id` and `score` in this order. Keep 1st row as header. For example: `q1 doc1 1`
358
+
359
+ ### Data Instances
360
+
361
+ A high level example of any beir dataset:
362
+
363
+ ```python
364
+ corpus = {
365
+ "doc1" : {
366
+ "title": "Albert Einstein",
367
+ "text": "Albert Einstein was a German-born theoretical physicist. who developed the theory of relativity, \
368
+ one of the two pillars of modern physics (alongside quantum mechanics). His work is also known for \
369
+ its influence on the philosophy of science. He is best known to the general public for his mass–energy \
370
+ equivalence formula E = mc2, which has been dubbed 'the world's most famous equation'. He received the 1921 \
371
+ Nobel Prize in Physics 'for his services to theoretical physics, and especially for his discovery of the law \
372
+ of the photoelectric effect', a pivotal step in the development of quantum theory."
373
+ },
374
+ "doc2" : {
375
+ "title": "", # Keep title an empty string if not present
376
+ "text": "Wheat beer is a top-fermented beer which is brewed with a large proportion of wheat relative to the amount of \
377
+ malted barley. The two main varieties are German Weißbier and Belgian witbier; other types include Lambic (made\
378
+ with wild yeast), Berliner Weisse (a cloudy, sour beer), and Gose (a sour, salty beer)."
379
+ },
380
+ }
381
+
382
+ queries = {
383
+ "q1" : "Who developed the mass-energy equivalence formula?",
384
+ "q2" : "Which beer is brewed with a large proportion of wheat?"
385
+ }
386
+
387
+ qrels = {
388
+ "q1" : {"doc1": 1},
389
+ "q2" : {"doc2": 1},
390
+ }
391
+ ```
392
+
393
+ ### Data Fields
394
+
395
+ Examples from all configurations have the following features:
396
+
397
+ ### Corpus
398
+ - `corpus`: a `dict` feature representing the document title and passage text, made up of:
399
+ - `_id`: a `string` feature representing the unique document id
400
+ - `title`: a `string` feature, denoting the title of the document.
401
+ - `text`: a `string` feature, denoting the text of the document.
402
+
403
+ ### Queries
404
+ - `queries`: a `dict` feature representing the query, made up of:
405
+ - `_id`: a `string` feature representing the unique query id
406
+ - `text`: a `string` feature, denoting the text of the query.
407
+
408
+ ### Qrels
409
+ - `qrels`: a `dict` feature representing the query document relevance judgements, made up of:
410
+ - `_id`: a `string` feature representing the query id
411
+ - `_id`: a `string` feature, denoting the document id.
412
+ - `score`: a `int32` feature, denoting the relevance judgement between query and document.
413
+
414
+
415
+ ### Data Splits
416
+
417
+ | Dataset | Website| BEIR-Name | Type | Queries | Corpus | Rel D/Q | Down-load | md5 |
418
+ | -------- | -----| ---------| --------- | ----------- | ---------| ---------| :----------: | :------:|
419
+ | MSMARCO | [Homepage](https://microsoft.github.io/msmarco/)| ``msmarco`` | ``train``<br>``dev``<br>``test``| 6,980 | 8.84M | 1.1 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/msmarco.zip) | ``444067daf65d982533ea17ebd59501e4`` |
420
+ | TREC-COVID | [Homepage](https://ir.nist.gov/covidSubmit/index.html)| ``trec-covid``| ``test``| 50| 171K| 493.5 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/trec-covid.zip) | ``ce62140cb23feb9becf6270d0d1fe6d1`` |
421
+ | NFCorpus | [Homepage](https://www.cl.uni-heidelberg.de/statnlpgroup/nfcorpus/) | ``nfcorpus`` | ``train``<br>``dev``<br>``test``| 323 | 3.6K | 38.2 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/nfcorpus.zip) | ``a89dba18a62ef92f7d323ec890a0d38d`` |
422
+ | BioASQ | [Homepage](http://bioasq.org) | ``bioasq``| ``train``<br>``test`` | 500 | 14.91M | 8.05 | No | [How to Reproduce?](https://github.com/UKPLab/beir/blob/main/examples/dataset#2-bioasq) |
423
+ | NQ | [Homepage](https://ai.google.com/research/NaturalQuestions) | ``nq``| ``train``<br>``test``| 3,452 | 2.68M | 1.2 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/nq.zip) | ``d4d3d2e48787a744b6f6e691ff534307`` |
424
+ | HotpotQA | [Homepage](https://hotpotqa.github.io) | ``hotpotqa``| ``train``<br>``dev``<br>``test``| 7,405 | 5.23M | 2.0 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/hotpotqa.zip) | ``f412724f78b0d91183a0e86805e16114`` |
425
+ | FiQA-2018 | [Homepage](https://sites.google.com/view/fiqa/) | ``fiqa`` | ``train``<br>``dev``<br>``test``| 648 | 57K | 2.6 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/fiqa.zip) | ``17918ed23cd04fb15047f73e6c3bd9d9`` |
426
+ | Signal-1M(RT) | [Homepage](https://research.signal-ai.com/datasets/signal1m-tweetir.html)| ``signal1m`` | ``test``| 97 | 2.86M | 19.6 | No | [How to Reproduce?](https://github.com/UKPLab/beir/blob/main/examples/dataset#4-signal-1m) |
427
+ | TREC-NEWS | [Homepage](https://trec.nist.gov/data/news2019.html) | ``trec-news`` | ``test``| 57 | 595K | 19.6 | No | [How to Reproduce?](https://github.com/UKPLab/beir/blob/main/examples/dataset#1-trec-news) |
428
+ | ArguAna | [Homepage](http://argumentation.bplaced.net/arguana/data) | ``arguana``| ``test`` | 1,406 | 8.67K | 1.0 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/arguana.zip) | ``8ad3e3c2a5867cdced806d6503f29b99`` |
429
+ | Touche-2020| [Homepage](https://webis.de/events/touche-20/shared-task-1.html) | ``webis-touche2020``| ``test``| 49 | 382K | 19.0 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/webis-touche2020.zip) | ``46f650ba5a527fc69e0a6521c5a23563`` |
430
+ | CQADupstack| [Homepage](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | ``cqadupstack``| ``test``| 13,145 | 457K | 1.4 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/cqadupstack.zip) | ``4e41456d7df8ee7760a7f866133bda78`` |
431
+ | Quora| [Homepage](https://www.quora.com/q/quoradata/First-Quora-Dataset-Release-Question-Pairs) | ``quora``| ``dev``<br>``test``| 10,000 | 523K | 1.6 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/quora.zip) | ``18fb154900ba42a600f84b839c173167`` |
432
+ | DBPedia | [Homepage](https://github.com/iai-group/DBpedia-Entity/) | ``dbpedia-entity``| ``dev``<br>``test``| 400 | 4.63M | 38.2 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/dbpedia-entity.zip) | ``c2a39eb420a3164af735795df012ac2c`` |
433
+ | SCIDOCS| [Homepage](https://allenai.org/data/scidocs) | ``scidocs``| ``test``| 1,000 | 25K | 4.9 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/scidocs.zip) | ``38121350fc3a4d2f48850f6aff52e4a9`` |
434
+ | FEVER | [Homepage](http://fever.ai) | ``fever``| ``train``<br>``dev``<br>``test``| 6,666 | 5.42M | 1.2| [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/fever.zip) | ``5a818580227bfb4b35bb6fa46d9b6c03`` |
435
+ | Climate-FEVER| [Homepage](http://climatefever.ai) | ``climate-fever``|``test``| 1,535 | 5.42M | 3.0 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/climate-fever.zip) | ``8b66f0a9126c521bae2bde127b4dc99d`` |
436
+ | SciFact| [Homepage](https://github.com/allenai/scifact) | ``scifact``| ``train``<br>``test``| 300 | 5K | 1.1 | [Link](https://public.ukp.informatik.tu-darmstadt.de/thakur/BEIR/datasets/scifact.zip) | ``5f7d1de60b170fc8027bb7898e2efca1`` |
437
+ | Robust04 | [Homepage](https://trec.nist.gov/data/robust/04.guidelines.html) | ``robust04``| ``test``| 249 | 528K | 69.9 | No | [How to Reproduce?](https://github.com/UKPLab/beir/blob/main/examples/dataset#3-robust04) |
438
+
439
+
440
+ ## Dataset Creation
441
+
442
+ ### Curation Rationale
443
+
444
+ [Needs More Information]
445
+
446
+ ### Source Data
447
+
448
+ #### Initial Data Collection and Normalization
449
+
450
+ [Needs More Information]
451
+
452
+ #### Who are the source language producers?
453
+
454
+ [Needs More Information]
455
+
456
+ ### Annotations
457
+
458
+ #### Annotation process
459
+
460
+ [Needs More Information]
461
+
462
+ #### Who are the annotators?
463
+
464
+ [Needs More Information]
465
+
466
+ ### Personal and Sensitive Information
467
+
468
+ [Needs More Information]
469
+
470
+ ## Considerations for Using the Data
471
+
472
+ ### Social Impact of Dataset
473
+
474
+ [Needs More Information]
475
+
476
+ ### Discussion of Biases
477
+
478
+ [Needs More Information]
479
+
480
+ ### Other Known Limitations
481
+
482
+ [Needs More Information]
483
+
484
+ ## Additional Information
485
+
486
+ ### Dataset Curators
487
+
488
+ [Needs More Information]
489
+
490
+ ### Licensing Information
491
+
492
+ [Needs More Information]
493
+
494
+ ### Citation Information
495
+
496
+ Cite as:
497
+ ```
498
+ @inproceedings{
499
+ thakur2021beir,
500
+ title={{BEIR}: A Heterogeneous Benchmark for Zero-shot Evaluation of Information Retrieval Models},
501
+ author={Nandan Thakur and Nils Reimers and Andreas R{\"u}ckl{\'e} and Abhishek Srivastava and Iryna Gurevych},
502
+ booktitle={Thirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 2)},
503
+ year={2021},
504
+ url={https://openreview.net/forum?id=wCu6T5xFjeJ}
505
+ }
506
+ ```
507
+
508
+ ### Contributions
509
+
510
+ Thanks to [@Nthakur20](https://github.com/Nthakur20) for adding this dataset.
train.jsonl.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f522bb4b8025f6573949a3972f538030480665d2bd75bf0c216e6771404a6173
3
+ size 6621873