Back to home
All Datasets 🏷
All datasets from our datasets repository and community bucket.
Also check out the list of supported Metrics 📉.
58 results
AdversarialQA is a Reading Comprehension dataset, consisting of questions posed by crowdworkers on a set of Wikipedia articles using an adversarial model-in-the-loop. We use three different models; BiDAF (Seo et al., 2016), BERT-Large (Devlin et al., 2018), and RoBERTa-Large (Liu et al., 2019) in the annotation loop and construct three datasets; D(BiDAF), D(BERT), and D(RoBERTa), each with 10,000 training examples, 1,000 validation, and 1,000 test examples. The adversarial human annotation paradigm ensures that these datasets consist of questions that current state-of-the-art models (at least the ones used as adversaries in the annotation loop) find challenging.
AmbigNQ, a dataset covering 14,042 questions from NQ-open, an existing open-domain QA benchmark. We find that over half of the questions in NQ-open are ambiguous. The types of ambiguity are diverse and sometimes subtle, many of which are only apparent after examining evidence provided by a very large text corpus. AMBIGNQ, a dataset with 14,042 annotations on NQ-OPEN questions containing diverse types of ambiguity. We provide two distributions of our new dataset AmbigNQ: a full version with all annotation metadata and a light version with only inputs and outputs.
A large-scale dataset consisting of approximately 100,000 algebraic word problems. The solution to each question is explained step-by-step using natural language. This data is used to train a program generation model that learns to generate the explanation, while generating the program that solves the question.
Machine reading comprehension tasks require a machine reader to answer questions relevant to the given document. In this paper, we present the first free-form multiple-Choice Chinese machine reading Comprehension dataset (C^3), containing 13,369 documents (dialogues or more formally written mixed-genre texts) and their associated 19,577 multiple-choice free-form questions collected from Chinese-as-a-second-language examinations. We present a comprehensive analysis of the prior knowledge (i.e., linguistic, domain-specific, and general world knowledge) needed for these real-world problems. We implement rule-based and popular neural methods and find that there is still a significant performance gap between the best performing model (68.5%) and human readers (96.0%), especially on problems that require prior knowledge. We further study the effects of distractor plausibility and data augmentation based on translated relevant datasets for English on model performance. We expect C^3 to present great challenges to existing systems as answering 86.8% of questions requires both knowledge within and beyond the accompanying document, and we hope that C^3 can serve as a platform to study how to leverage various kinds of prior knowledge to better understand a given written or orally oriented text.
The COmmonsense Dataset Adversarially-authored by Humans (CODAH) is an evaluation set for commonsense question-answering in the sentence completion style of SWAG. As opposed to other automatically generated NLI datasets, CODAH is adversarially constructed by humans who can view feedback from a pre-trained model and use this information to design challenging commonsense questions. Our experimental results show that CODAH questions present a complementary extension to the SWAG dataset, testing additional modes of common sense.
Doc2dial is dataset of goal-oriented dialogues that are grounded in the associated documents. It includes over 4500 annotated conversations with an average of 14 turns that are grounded in over 450 documents from four domains. Compared to the prior document-grounded dialogue datasets this dataset covers a variety of dialogue scenes in information-seeking conversations.
DREAM is a multiple-choice Dialogue-based REAding comprehension exaMination dataset. In contrast to existing reading comprehension datasets, DREAM is the first to focus on in-depth multi-turn multi-party dialogue understanding.
The Did You Know (pol. Czy wiesz?) dataset consists of human-annotated question-answer pairs. The task is to predict if the answer is correct. We chose the negatives which have the largest token overlap with a question.
EXAMS is a benchmark dataset for multilingual and cross-lingual question answering from high school examinations. It consists of more than 24,000 high-quality high school exam questions in 16 languages, covering 8 language families and 24 school subjects from Natural Sciences and Social Sciences, among others.
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.
HEAD-QA is a multi-choice HEAlthcare Dataset. The questions come from exams to access a specialized position in the Spanish healthcare system, and are challenging even for highly specialized humans. They are designed by the Ministerio de Sanidad, Consumo y Bienestar Social. The dataset contains questions about the following topics: medicine, nursing, psychology, chemistry, pharmacology and biology.
Existing question answering datasets focus on dealing with homogeneous information, based either only on text or KB/Table information alone. However, as human knowledge is distributed over heterogeneous forms, using homogeneous information alone might lead to severe coverage problems. To fill in the gap, we present HybridQA, a new large-scale question-answering dataset that requires reasoning on heterogeneous information. Each question is aligned with a Wikipedia table and multiple free-form corpora linked with the entities in the table. The questions are designed to aggregate both tabular information and text information, i.e., lack of either form would render the question unanswerable.
KILT tasks training and evaluation data. - [FEVER](https://fever.ai) | Fact Checking | fever - [AIDA CoNLL-YAGO](https://www.mpi-inf.mpg.de/departments/databases-and-information-systems/research/ambiverse-nlu/aida/downloads) | Entity Linking | aidayago2 - [WNED-WIKI](https://github.com/U-Alberta/wned) | Entity Linking | wned - [WNED-CWEB](https://github.com/U-Alberta/wned) | Entity Linking | cweb - [T-REx](https://hadyelsahar.github.io/t-rex) | Slot Filling | trex - [Zero-Shot RE](http://nlp.cs.washington.edu/zeroshot) | Slot Filling | structured_zeroshot - [Natural Questions](https://ai.google.com/research/NaturalQuestions) | Open Domain QA | nq - [HotpotQA](https://hotpotqa.github.io) | Open Domain QA | hotpotqa - [TriviaQA](http://nlp.cs.washington.edu/triviaqa) | Open Domain QA | triviaqa - [ELI5](https://facebookresearch.github.io/ELI5/explore.html) | Open Domain QA | eli5 - [Wizard of Wikipedia](https://parl.ai/projects/wizard_of_wikipedia) | Dialogue | wow To finish linking TriviaQA questions to the IDs provided, follow the instructions [here](http://github.com/huggingface/datasets/datasets/kilt_tasks/README.md).
This is LiveQA, a Chinese dataset constructed from play-by-play live broadcast. It contains 117k multiple-choice questions written by human commentators for over 1,670 NBA games, which are collected from the Chinese Hupu website.
MC-TACO (Multiple Choice TemporAl COmmonsense) is a dataset of 13k question-answer pairs that require temporal commonsense comprehension. A system receives a sentence providing context information, a question designed to require temporal commonsense knowledge, and multiple candidate answers. More than one candidate answer can be plausible. The task is framed as binary classification: givent he context, the question, and the candidate answer, the task is to determine whether the candidate answer is plausible ("yes") or not ("no").
MedHop is based on research paper abstracts from PubMed, and the queries are about interactions between pairs of drugs. The correct answer has to be inferred by combining information from a chain of reactions of drugs and proteins.
The MedDialog dataset (English) contains conversations (in English) between doctors and patients.It has 0.26 million dialogues. The data is continuously growing and more dialogues will be added. The raw dialogues are from healthcaremagic.com and icliniq.com. All copyrights of the data belong to healthcaremagic.com and icliniq.com.
We introduce MKQA, an open-domain question answering evaluation set comprising 10k question-answer pairs sampled from the Google Natural Questions dataset, aligned across 26 typologically diverse languages (260k question-answer pairs in total). For each query we collected new passage-independent answers. These queries and answers were then human translated into 25 Non-English languages.
Posing reading comprehension as a generation problem provides a great deal of flexibility, allowing for open-ended questions with few restrictions on possible answers. However, progress is impeded by existing generation metrics, which rely on token overlap and are agnostic to the nuances of reading comprehension. To address this, we introduce a benchmark for training and evaluating generative reading comprehension metrics: MOdeling Correctness with Human Annotations. MOCHA contains 40K human judgement scores on model outputs from 6 diverse question answering datasets and an additional set of minimal pairs for evaluation. Using MOCHA, we train an evaluation metric: LERC, a Learned Evaluation metric for Reading Comprehension, to mimic human judgement scores.
The MRQA 2019 Shared Task focuses on generalization in question answering. An effective question answering system should do more than merely interpolate from the training set to answer test examples drawn from the same distribution: it should also be able to extrapolate to out-of-distribution examples — a significantly harder challenge. The dataset is a collection of 18 existing QA dataset (carefully selected subset of them) and converted to the same format (SQuAD format). Among these 18 datasets, six datasets were made available for training, six datasets were made available for development, and the final six for testing. The dataset is released as part of the MRQA 2019 Shared Task.
Recent work in semantic parsing for question answering has focused on long and complicated questions, many of which would seem unnatural if asked in a normal conversation between two humans. In an effort to explore a conversational QA setting, we present a more realistic task: answering sequences of simple but inter-related questions. We created SQA by asking crowdsourced workers to decompose 2,022 questions from WikiTableQuestions (WTQ), which contains highly-compositional questions about tables from Wikipedia. We had three workers decompose each WTQ question, resulting in a dataset of 6,066 sequences that contain 17,553 questions in total. Each question is also associated with answers in the form of cell locations in the tables.
MultiReQA contains the sentence boundary annotation from eight publicly available QA datasets including SearchQA, TriviaQA, HotpotQA, NaturalQuestions, SQuAD, BioASQ, RelationExtraction, and TextbookQA. Five of these datasets, including SearchQA, TriviaQA, HotpotQA, NaturalQuestions, SQuAD, contain both training and test data, and three, including BioASQ, RelationExtraction, TextbookQA, contain only the test data
The NarrativeQA dataset for question answering on long documents (movie scripts, books). It includes the list of documents with Wikipedia summaries, links to full stories, and questions and answers.
Neural-Code-Search-Evaluation-Dataset presents an evaluation dataset consisting of natural language query and code snippet pairs and a search corpus consisting of code snippets collected from the most popular Android repositories on GitHub.
NewsQA is a challenging machine comprehension dataset of over 100,000 human-generated question-answer pairs. Crowdworkers supply questions and answers based on a set of over 10,000 news articles from CNN, with answers consisting of spans of text from the corresponding articles.
The NQ-Open task, introduced by Lee et.al. 2019, is an open domain question answering benchmark that is derived from Natural Questions. The goal is to predict an English answer string for an input English question. All questions can be answered using the contents of English Wikipedia.
To apply eyeshadow without a brush, should I use a cotton swab or a toothpick? Questions requiring this kind of physical commonsense pose a challenge to state-of-the-art natural language understanding systems. The PIQA dataset introduces the task of physical commonsense reasoning and a corresponding benchmark dataset Physical Interaction: Question Answering or PIQA. Physical commonsense knowledge is a major challenge on the road to true AI-completeness, including robots that interact with the world and understand natural language. PIQA focuses on everyday situations with a preference for atypical solutions. The dataset is inspired by instructables.com, which provides users with instructions on how to build, craft, bake, or manipulate objects using everyday materials. The underlying task is formualted as multiple choice question answering: given a question `q` and two possible solutions `s1`, `s2`, a model or a human must choose the most appropriate solution, of which exactly one is correct. The dataset is further cleaned of basic artifacts using the AFLite algorithm which is an improvement of adversarial filtering. The dataset contains 16,000 examples for training, 2,000 for development and 3,000 for testing.
This dataset is for studying computational models trained to reason about prototypical situations. Using deterministic filtering a sampling from a larger set of all transcriptions was built. It contains 9789 instances where each instance represents a survey question from Family Feud game. Each instance exactly is a question, a set of answers, and a count associated with each answer. Each line is a json dictionary, in which: 1. question - contains the question (in original and a normalized form) 2. answerstrings - contains the original answers provided by survey respondents (when available), along with the counts for each string. Because the FamilyFeud data has only cluster names rather than strings, those cluster names are included with 0 weight. 3. answer-clusters - lists clusters, with the count of each cluster and the strings included in that cluster. Each cluster is given a unique ID that can be linked to in the assessment files.
PubMedQA is a novel biomedical question answering (QA) dataset collected from PubMed abstracts. The task of PubMedQA is to answer research questions with yes/no/maybe (e.g.: Do preoperative statins reduce atrial fibrillation after coronary artery bypass grafting?) using the corresponding abstracts. PubMedQA has 1k expert-annotated, 61.2k unlabeled and 211.3k artificially generated QA instances. Each PubMedQA instance is composed of (1) a question which is either an existing research article title or derived from one, (2) a context which is the corresponding abstract without its conclusion, (3) a long answer, which is the conclusion of the abstract and, presumably, answers the research question, and (4) a yes/no/maybe answer which summarizes the conclusion. PubMedQA is the first QA dataset where reasoning over biomedical research texts, especially their quantitative contents, is required to answer the questions.
The dataset contains question-answer pairs to model verbal predicate-argument structure. The questions start with wh-words (Who, What, Where, What, etc.) and contain a verb predicate in the sentence; the answers are phrases in the sentence. There were 2 datsets used in the paper, newswire and wikipedia. Unfortunately the newswiredataset is built from CoNLL-2009 English training set that is covered under license Thus, we are providing only Wikipedia training set here. Please check README.md for more details on newswire dataset. For the Wikipedia domain, randomly sampled sentences from the English Wikipedia (excluding questions and sentences with fewer than 10 or more than 60 words) were taken. This new dataset is designed to solve this great NLP task and is crafted with a lot of care.
QED, is a linguistically informed, extensible framework for explanations in question answering. A QED explanation specifies the relationship between a question and answer according to formal semantic notions such as referential equality, sentencehood, and entailment. It is an expertannotated dataset of QED explanations built upon a subset of the Google Natural Questions dataset.
Question Answering in Context is a dataset for modeling, understanding, and participating in information seeking dialog. Data instances consist of an interactive dialog between two crowd workers: (1) a student who poses a sequence of freeform questions to learn as much as possible about a hidden Wikipedia text, and (2) a teacher who answers the questions by providing short excerpts (spans) from the text. QuAC introduces challenges not found in existing machine comprehension datasets: its questions are often more open-ended, unanswerable, or only meaningful within the dialog context.
ROPES (Reasoning Over Paragraph Effects in Situations) is a QA dataset which tests a system's ability to apply knowledge from a passage of text to a new situation. A system is presented a background passage containing a causal or qualitative relation(s) (e.g., "animal pollinators increase efficiency of fertilization in flowers"), a novel situation that uses this background, and questions that require reasoning about effects of the relationships in the background passage in the background of the situation.
ShARC is a Conversational Question Answering dataset focussing on question answering from texts containing rules. The goal is to answer questions by possibly asking follow-up questions first. It is assumed assume that the question is often underspecified, in the sense that the question does not provide enough information to be answered directly. However, an agent can use the supporting rule text to infer what needs to be asked in order to determine the final answer.
ShARC, a conversational QA task, requires a system to answer user questions based on rules expressed in natural language text. However, it is found that in the ShARC dataset there are multiple spurious patterns that could be exploited by neural models. SharcModified is a new dataset which reduces the patterns identified in the original dataset. To reduce the sensitivity of neural models, for each occurence of an instance conforming to any of the patterns, we automatically construct alternatives where we choose to either replace the current instance with an alternative instance which does not exhibit the pattern; or retain the original instance. The modified ShARC has two versions sharc-mod and history-shuffled. For morre details refer to Appendix A.3 .
SimpleQuestions is a dataset for simple QA, which consists of a total of 108,442 questions written in natural language by human English-speaking annotators each paired with a corresponding fact, formatted as (subject, relationship, object), that provides the answer but also a complete explanation. Fast have been extracted from the Knowledge Base Freebase (freebase.com). We randomly shuffle these questions and use 70% of them (75910) as training set, 10% as validation set (10845), and the remaining 20% as test set.
Dataset with the text of 10% of questions and answers from the Stack Overflow programming Q&A website. This is organized as three tables: Questions contains the title, body, creation date, closed date (if applicable), score, and owner ID for all non-deleted Stack Overflow questions whose Id is a multiple of 10. Answers contains the body, creation date, score, and owner ID for each of the answers to these questions. The ParentId column links back to the Questions table. Tags contains the tags on each of these questions.
Here are two different adversaries, each of which uses a different procedure to pick the sentence it adds to the paragraph: AddSent: Generates up to five candidate adversarial sentences that don't answer the question, but have a lot of words in common with the question. Picks the one that most confuses the model. AddOneSent: Similar to AddSent, but just picks one of the candidate sentences at random. This adversary is does not query the model in any way.
KorQuAD 1.0 is a large-scale Korean dataset for machine reading comprehension task consisting of human generated questions for Wikipedia articles. We benchmark the data collecting process of SQuADv1.0 and crowdsourced 70,000+ question-answer pairs. 1,637 articles and 70,079 pairs of question answers were collected. 1,420 articles are used for the training set, 140 for the dev set, and 77 for the test set. 60,407 question-answer pairs are for the training set, 5,774 for the dev set, and 3,898 for the test set.
KorQuAD 2.0 is a Korean question and answering dataset consisting of a total of 100,000+ pairs. There are three major differences from KorQuAD 1.0, which is the standard Korean Q & A data. The first is that a given document is a whole Wikipedia page, not just one or two paragraphs. Second, because the document also contains tables and lists, it is necessary to understand the document structured with HTML tags. Finally, the answer can be a long text covering not only word or phrase units, but paragraphs, tables, and lists. As a baseline model, BERT Multilingual is used, released by Google as an open source. It shows 46.0% F1 score, a very low score compared to 85.7% of the human F1 score. It indicates that this data is a challenging task. Additionally, we increased the performance by no-answer data augmentation. Through the distribution of this data, we intend to extend the limit of MRC that was limited to plain text to real world tasks of various lengths and formats.
`thaiqa_squad` is an open-domain, extractive question answering dataset (4,000 questions in `train` and 74 questions in `dev`) in [SQuAD](https://rajpurkar.github.io/SQuAD-explorer/) format, originally created by [NECTEC](https://www.nectec.or.th/en/) from Wikipedia articles and adapted to [SQuAD](https://rajpurkar.github.io/SQuAD-explorer/) format by [PyThaiNLP](https://github.com/PyThaiNLP/).
WikiHop is open-domain and based on Wikipedia articles; the goal is to recover Wikidata information by hopping through documents. The goal is to answer text understanding queries by combining multiple facts that are spread across different documents.
\ The dataset extracted from Persian Wikipedia into the form of articles and highlights and cleaned the dataset into pairs of articles and highlights and reduced the articles' length (only version 1.0.0) and highlights' length to a maximum of 512 and 128, respectively, suitable for parsBERT.
XGLUE is a new benchmark dataset to evaluate the performance of cross-lingual pre-trained models with respect to cross-lingual natural language understanding and generation. The benchmark is composed of the following 11 tasks: - NER - POS Tagging (POS) - News Classification (NC) - MLQA - XNLI - PAWS-X - Query-Ad Matching (QADSM) - Web Page Ranking (WPR) - QA Matching (QAM) - Question Generation (QG) - News Title Generation (NTG) For more information, please take a look at https://microsoft.github.io/XGLUE/.
XOR-TyDi QA brings together for the first time information-seeking questions, open-retrieval QA, and multilingual QA to create a multilingual open-retrieval QA dataset that enables cross-lingual answer retrieval. It consists of questions written by information-seeking native speakers in 7 typologically diverse languages and answer annotations that are retrieved from multilingual document collections. There are three sub-tasks: XOR-Retrieve, XOR-EnglishSpan, and XOR-Full.
XQuAD-R is a retrieval version of the XQuAD dataset (a cross-lingual extractive QA dataset). Like XQuAD, XQUAD-R is an 11-way parallel dataset, where each question appears in 11 different languages and has 11 parallel correct answers across the languages.
Yahoo Non-Factoid Question Dataset is derived from Yahoo's Webscope L6 collection using machine learning techiques such that the questions would contain non-factoid answers.The dataset contains 87,361 questions and their corresponding answers. Each question contains its best answer along with additional other answers submitted by users. Only the best answer was reviewed in determining the quality of the question-answer pair.
ZEST tests whether NLP systems can perform unseen tasks in a zero-shot way, given a natural language description of the task. It is an instantiation of our proposed framework "learning from task descriptions". The tasks include classification, typed entity extraction and relationship extraction, and each task is paired with 20 different annotated (input, output) examples. ZEST's structure allows us to systematically test whether models can generalize in five different ways.