GrailQA

The Strongly Generalizable Question Answering Dataset

What is GrailQA?

Strongly Generalizable Question Answering (GrailQA) is a new large-scale, high-quality dataset for question answering on knowledge bases (KBQA) on Freebase with 64,331 questions annotated with both answers and corresponding logical forms in different syntax (i.e., SPARQL, S-expression, etc.). It can be used to test three levels of generalization in KBQA: i.i.d., compositional, and zero-shot.

Explore GrailQAGrailQA paper (Gu et al. '20)Code

Why GrailQA?

GrailQA is by far the largest crowdsourced KBQA dataset with questions of high diversity (i.e., questions in GrailQA can have up to 4 relations and optionally have a function from counting, superlatives and comparatives). It also has the highest coverage over Freebase; it widely covers 3,720 relations and 86 domains from Freebase. Last but not least, our meticulous data split allows GrailQA to test not only i.i.d. generalization, but also compositional generalization and zero-shot generalization, which are critical for practical KBQA systems.

News

  • 01/24/2021We provide instructions on Freebase setup.
  • 01/22/2021We have updated our baseline performance based on a new entity linker with a recall of around 0.77 on the dev set, compared with the previous recall of around 0.46. More details about entity linking can be found in the updated paper.
  • 01/18/2021 We will have some major update on our baseline model results with a new entity linker in this week. The numbers will be higher. Please stay tuned!
  • 11/30/2020 We fix some minor error in the sparql_queries provided in our dataset.

Getting Started

We've built a few resources to help you get started with the dataset.

Download a copy of the dataset (distributed under the CC BY-SA 4.0 license):


To work with our dataset, we recommend you setting up a Virtuoso server for Freebase (feel free to choose your own way to index Freebase). Please find both a clean version of Freebase dump and instructions on setting up the server via:


To evaluate your models, we have also made available the evaluation script we will use for official evaluation, along with a sample prediction file that the script will take as input. Evaluating semantic-level exact match also depends on several preprocessed ontology files of Freebase. You can find all of them here. To run the evaluation, use python evaluate.py <path_to_dev> <path_to_predictions> --fb_roles <path_to_fb_roles> --fb_types <path_to_fb_types> --reverse_properties <path_to_reverse_properties>.

Once you have a built a model that works to your expectations on the dev set, you submit it to get official scores on the dev and a hidden test set. To preserve the integrity of test results, we do not release the labels of test set to the public. Here's a tutorial walking you through official evaluation of your model:

Submission Tutorial

Have Questions?

Send an email to gu.826@osu.edu, or create an issue in github.

Acknowledgement

We thank Pranav Rajpurkar and Robin Jia for giving us the permission to build this website based on SQuAD.

Leaderboard: Overall

Here are the overall Exact Match (EM) and F1 scores evaluated on GrailQA test set. To get the EM score on GrailQA, please submit your results with logical forms in S-expression. Note that, submissions are ranked only based on F1, so feel free to choose your own meaning representation as EM won't affect your ranking.

RankModelEMF1

1

Aug 20, 2021
RnG-KBQA (single model)

Salesforce Research

68.77874.422

2

Aug 12, 2021
S2QL (single model)

Anonymous

57.45666.186

3

Apr 05, 2021
ReTraCk (single model)

Microsoft Research Asia

https://aclanthology.org/2021.acl-demo.39/
58.13665.285

4

Feb 04, 2021
ArcaneQA (single model)

Anonymous

57.87264.924

5

Jan 22, 2021
BERT+Ranking (single model)

The Ohio State University

50.57857.988

6

Jan 22, 2021
GloVe+Ranking (single model)

The Ohio State University

39.52145.136

7

Jan 22, 2021
BERT+Transduction (single model)

The Ohio State University

33.25536.803

8

Jan 22, 2021
GloVe+Transduction (single model)

The Ohio State University

17.58718.432

Leaderboard: Compositional Generalization

Here are the Exact Match (EM) and F1 scores evaluated on the subset of GrailQA test set that tests compositional generalization.

RankModelEMF1

1

Aug 20, 2021
RnG-KBQA (single model)

Salesforce Research

63.79271.156

2

Apr 05, 2021
ReTraCk (single model)

Microsoft Research Asia

https://aclanthology.org/2021.acl-demo.39/
61.49970.911

3

Aug 12, 2021
S2QL (single model)

Anonymous

54.71664.679

4

Feb 04, 2021
ArcaneQA (single model)

Anonymous

56.39563.533

5

Jan 22, 2021
BERT+Ranking (single model)

The Ohio State University

45.51053.890

6

Jan 22, 2021
GloVe+Ranking (single model)

The Ohio State University

39.95547.753

7

Jan 22, 2021
BERT+Transduction (single model)

The Ohio State University

31.04035.985

8

Jan 22, 2021
GloVe+Transduction (single model)

The Ohio State University

16.44118.507

Leaderboard: Zero-shot Generalization

Here are the Exact Match (EM) and F1 scores evaluated on the subset of GrailQA test set that tests zero-shot generalization.

RankModelEMF1

1

Aug 20, 2021
RnG-KBQA (single model)

Salesforce Research

62.98869.182

2

Aug 12, 2021
S2QL (single model)

Anonymous

55.12263.598

3

Feb 04, 2021
ArcaneQA (single model)

Anonymous

49.96458.844

4

Jan 22, 2021
BERT+Ranking (single model)

The Ohio State University

48.56655.660

5

Apr 05, 2021
ReTraCk (single model)

Microsoft Research Asia

https://aclanthology.org/2021.acl-demo.39/
44.56152.539

6

Jan 22, 2021
GloVe+Ranking (single model)

The Ohio State University

28.88633.792

7

Jan 22, 2021
BERT+Transduction (single model)

The Ohio State University

25.70229.300

8

Jan 22, 2021
GloVe+Transduction (single model)

The Ohio State University

2.9683.123