List of datasets for machine-learning research

These datasets are used for machine-learning research and have been cited in peer-reviewed academic journals. Datasets are an integral part of the field of machine learning. Major advances in this field can result from advances in learning algorithms (such as deep learning), computer hardware, and, less-intuitively, the availability of high-quality training datasets. High-quality labeled training datasets for supervised and semi-supervised machine learning algorithms are usually difficult and expensive to produce because of the large amount of time needed to label the data. Although they do not need to be labeled, high-quality datasets for unsupervised learning can also be difficult and costly to produce.

Image data

Datasets consisting primarily of images or videos for tasks such as object detection, facial recognition, and multi-label classification.

Facial recognition

In computer vision, face images have been used extensively to develop facial recognition systems, face detection, and many other projects that use images of faces.

Dataset name	Brief description	Preprocessing	Instances	Format	Default task	Created (updated)	Reference	Creator
Aff-Wild	298 videos of 200 individuals, ~1,250,000 manually annotated images: annotated in terms of dimensional affect (valence-arousal); in-the-wild setting; color database; various resolutions (average = 640x360)	the detected faces, facial landmarks and valence-arousal annotations	~1,250,000 manually annotated images	video (visual + audio modalities)	affect recognition (valence-arousal estimation)	2017	CVPR IJCV	D.Kollias et al.
Aff-Wild2	558 videos of 458 individuals, ~2,800,000 manually annotated images: annotated in terms of i) categorical affect (7 basic expressions: neutral, happiness, sadness, surprise, fear, disgust, anger); ii) dimensional affect (valence-arousal); iii) action units (AUs 1,2,4,6,12,15,20,25); in-the-wild setting; color database; various resolutions (average = 1030x630)	the detected faces, detected and aligned faces and annotations	~2,800,000 manually annotated images	video (visual + audio modalities)	affect recognition (valence-arousal estimation, basic expression classification, action unit detection)	2019	BMVC FG	D.Kollias et al.
FERET (facial recognition technology)	11338 images of 1199 individuals in different positions and at different times.	None.	11,338	Images	Classification, face recognition	2003		United States Department of Defense
Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS)	7,356 video and audio recordings of 24 professional actors. 8 emotions each at two intensities.	Files labelled with expression. Perceptual validation ratings provided by 319 raters.	7,356	Video, sound files	Classification, face recognition, voice recognition	2018		S.R. Livingstone and F.A. Russo
SCFace	Color images of faces at various angles.	Location of facial features extracted. Coordinates of features given.	4,160	Images, text	Classification, face recognition	2011		M. Grgic et al.
Yale Face Database	Faces of 15 individuals in 11 different expressions.	Labels of expressions.	165	Images	Face recognition	1997		J. Yang et al.
Cohn-Kanade AU-Coded Expression Database	Large database of images with labels for expressions.	Tracking of certain facial features.	500+ sequences	Images, text	Facial expression analysis	2000		T. Kanade et al.
JAFFE Facial Expression Database	213 images of 7 facial expressions (6 basic facial expressions + 1 neutral) posed by 10 Japanese female models.	Images are cropped to the facial region. Includes semantic ratings data on emotion labels.	213	Images, text	Facial expression cognition	1998		Lyons, Kamachi, Gyoba
FaceScrub	Images of public figures scrubbed from image searching.	Name and m/f annotation.	107,818	Images, text	Face recognition	2014		H. Ng et al.
BioID Face Database	Images of faces with eye positions marked.	Manually set eye positions.	1521	Images, text	Face recognition	2001		BioID
Skin Segmentation Dataset	Randomly sampled color values from face images.	B, G, R, values extracted.	245,057	Text	Segmentation, classification	2012		R. Bhatt.
Bosphorus	3D Face image database.	34 action units and 6 expressions labeled; 24 facial landmarks labeled.	4652	Images, text	Face recognition, classification	2008		A Savran et al.
UOY 3D-Face	neutral face, 5 expressions: anger, happiness, sadness, eyes closed, eyebrows raised.	labeling.	5250	Images, text	Face recognition, classification	2004		University of York
CASIA 3D Face Database	Expressions: Anger, smile, laugh, surprise, closed eyes.	None.	4624	Images, text	Face recognition, classification	2007		Institute of Automation, Chinese Academy of Sciences
CASIA NIR	Expressions: Anger Disgust Fear Happiness Sadness Surprise	None.	480	Annotated Visible Spectrum and Near Infrared Video captures at 25 frames per second	Face recognition, classification	2011		Zhao, G. et al.
BU-3DFE	neutral face, and 6 expressions: anger, happiness, sadness, surprise, disgust, fear (4 levels). 3D images extracted.	None.	2500	Images, text	Facial expression recognition, classification	2006		Binghamton University
Face Recognition Grand Challenge Dataset	Up to 22 samples for each subject. Expressions: anger, happiness, sadness, surprise, disgust, puffy. 3D Data.	None.	4007	Images, text	Face recognition, classification	2004		National Institute of Standards and Technology
Gavabdb	Up to 61 samples for each subject. Expressions neutral face, smile, frontal accentuated laugh, frontal random gesture. 3D images.	None.	549	Images, text	Face recognition, classification	2008		King Juan Carlos University
3D-RMA	Up to 100 subjects, expressions mostly neutral. Several poses as well.	None.	9971	Images, text	Face recognition, classification	2004		Royal Military Academy (Belgium)
SoF	112 persons (66 males and 46 females) wear glasses under different illumination conditions.	A set of synthetic filters (blur, occlusions, noise, and posterization ) with different level of difficulty.	42,592 (2,662 original image × 16 synthetic image)	Images, Mat file	Gender classification, face detection, face recognition, age estimation, and glasses detection	2017		Afifi, M. et al.
IMDB-WIKI	IMDB and Wikipedia face images with gender and age labels.	None	523,051	Images	Gender classification, face detection, face recognition, age estimation	2015		R. Rothe, R. Timofte, L. V. Gool

Action recognition

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
TV Human Interaction Dataset	Videos from 20 different TV shows for prediction social actions: handshake, high five, hug, kiss and none.	None.	6,766 video clips	video clips	Action prediction	2013		Patron-Perez, A. et al.
Berkeley Multimodal Human Action Database (MHAD)	Recordings of a single person performing 12 actions	MoCap pre-processing	660 action samples	8 PhaseSpace Motion Capture, 2 Stereo Cameras, 4 Quad Cameras, 6 accelerometers, 4 microphones	Action classification	2013		Ofli, F. et al.
THUMOS Dataset	Large video dataset for action classification.	Actions classified and labeled.	45M frames of video	Video, images, text	Classification, action detection	2013		Y. Jiang et al.
MEXAction2	Video dataset for action localization and spotting	Actions classified and labeled.	1000	Video	Action detection	2014		Stoian et al.

Object detection and recognition

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Visual Genome	Images and their description		108,000	images, text	Image captioning	2016		R. Krishna et al.
Berkeley 3-D Object Dataset	849 images taken in 75 different scenes. About 50 different object classes are labeled.	Object bounding boxes and labeling.	849	labeled images, text	Object recognition	2014		A. Janoch et al.
Berkeley Segmentation Data Set and Benchmarks 500 (BSDS500)	500 natural images, explicitly separated into disjoint train, validation and test subsets + benchmarking code. Based on BSDS300.	Each image segmented by five different subjects on average.	500	Segmented images	Contour detection and hierarchical image segmentation	2011		University of California, Berkeley
Microsoft Common Objects in Context (COCO)	complex everyday scenes of common objects in their natural context.	Object highlighting, labeling, and classification into 91 object types.	2,500,000	Labeled images, text	Object recognition	2015		T. Lin et al.
SUN Database	Very large scene and object recognition database.	Places and objects are labeled. Objects are segmented.	131,067	Images, text	Object recognition, scene recognition	2014		J. Xiao et al.
ImageNet	Labeled object image database, used in the ImageNet Large Scale Visual Recognition Challenge	Labeled objects, bounding boxes, descriptive words, SIFT features	14,197,122	Images, text	Object recognition, scene recognition	2009 (2014)		J. Deng et al.
Open Images	A Large set of images listed as having CC BY 2.0 license with image-level labels and bounding boxes spanning thousands of classes.	Image-level labels, Bounding boxes	9,178,275	Images, text	Classification, Object recognition	2017
TV News Channel Commercial Detection Dataset	TV commercials and news broadcasts.	Audio and video features extracted from still images.	129,685	Text	Clustering, classification	2015		P. Guha et al.
Statlog (Image Segmentation) Dataset	The instances were drawn randomly from a database of 7 outdoor images and hand-segmented to create a classification for every pixel.	Many features calculated.	2310	Text	Classification	1990		University of Massachusetts
Caltech 101	Pictures of objects.	Detailed object outlines marked.	9146	Images	Classification, object recognition.	2003		F. Li et al.
Caltech-256	Large dataset of images for object classification.	Images categorized and hand-sorted.	30,607	Images, Text	Classification, object detection	2007		G. Griffin et al.
SIFT10M Dataset	SIFT features of Caltech-256 dataset.	Extensive SIFT feature extraction.	11,164,866	Text	Classification, object detection	2016		X. Fu et al.
LabelMe	Annotated pictures of scenes.	Objects outlined.	187,240	Images, text	Classification, object detection	2005		MIT Computer Science and Artificial Intelligence Laboratory
Cityscapes Dataset	Stereo video sequences recorded in street scenes, with pixel-level annotations. Metadata also included.	Pixel-level segmentation and labeling	25,000	Images, text	Classification, object detection	2016		Daimler AG et al.
PASCAL VOC Dataset	Large number of images for classification tasks.	Labeling, bounding box included	500,000	Images, text	Classification, object detection	2010		M. Everingham et al.
CIFAR-10 Dataset	Many small, low-resolution, images of 10 classes of objects.	Classes labelled, training set splits created.	60,000	Images	Classification	2009		A. Krizhevsky et al.
CIFAR-100 Dataset	Like CIFAR-10, above, but 100 classes of objects are given.	Classes labelled, training set splits created.	60,000	Images	Classification	2009		A. Krizhevsky et al.
CINIC-10 Dataset	A unified contribution of CIFAR-10 and Imagenet with 10 classes, and 3 splits. Larger than CIFAR-10.	Classes labelled, training, validation, test set splits created.	270,000	Images	Classification	2018		Luke N. Darlow, Elliot J. Crowley, Antreas Antoniou, Amos J. Storkey
Fashion-MNIST	A MNIST-like fashion product database	Classes labelled, training set splits created.	60,000	Images	Classification	2017		Zalando SE
notMNIST	Some publicly available fonts and extracted glyphs from them to make a dataset similar to MNIST. There are 10 classes, with letters A-J taken from different fonts.	Classes labelled, training set splits created.	500,000	Images	Classification	2011		Yaroslav Bulatov
German Traffic Sign Detection Benchmark Dataset	Images from vehicles of traffic signs on German roads. These signs comply with UN standards and therefore are the same as in other countries.	Signs manually labeled	900	Images	Classification	2013		S Houben et al.
KITTI Vision Benchmark Dataset	Autonomous vehicles driving through a mid-size city captured images of various areas using cameras and laser scanners.	Many benchmarks extracted from data.	>100 GB of data	Images, text	Classification, object detection	2012		A Geiger et al.
Linnaeus 5 dataset	Images of 5 classes of objects.	Classes labelled, training set splits created.	8000	Images	Classification	2017		Chaladze & Kalatozishvili
FieldSAFE	Multi-modal dataset for obstacle detection in agriculture including stereo camera, thermal camera, web camera, 360-degree camera, lidar, radar, and precise localization.	Classes labelled geographically.	>400 GB of data	Images and 3D point clouds	Classification, object detection, object localization	2017		M. Kragh et al.
11K Hands	11,076 hand images (1600 x 1200 pixels) of 190 subjects, of varying ages between 18 – 75 years old, for gender recognition and biometric identification.	None	11,076 hand images	Images and (.mat, .txt, and .csv) label files	Gender recognition and biometric identification	2017		M Afifi
CORe50	Specifically designed for Continuous/Lifelong Learning and Object Recognition, is a collection of more than 500 videos (30fps) of 50 domestic objects belonging to 10 different categories.	Classes labelled, training set splits created based on a 3-way, multi-runs benchmark.	164,866 RBG-D images	images (.png or .pkl) and (.pkl, .txt, .tsv) label files	Classification, Object recognition	2017		V. Lomonaco and D. Maltoni
OpenLORIS-Object	Lifelong/Continual Robotic Vision dataset (OpenLORIS-Object) collected by real robots mounted with multiple high-resolution sensors, includes a collection of 121 object instances (1st version of dataset, 40 categories daily necessities objects under 20 scenes). The dataset has rigorously considered 4 environment factors under different scenes, including illumination, occlusion, object pixel size and clutter, and defines the difficulty levels of each factor explicitly.	Classes labelled, training/validation/testing set splits created by benchmark scripts.	1,106,424 RBG-D images	images (.png and .pkl) and (.pkl) label files	Classification, Lifelong object recognition, Robotic Vision	2019		Q. She et al.
THz and thermal video data set	This multispectral data set includes terahertz, thermal, visual, near infrared, and three-dimensional videos of objects hidden under people's clothes.	3D lookup tables are provided that allow you to project images onto 3D point clouds.	More than 20 videos. The duration of each video is about 85 seconds (about 345 frames).	AP2J	Experiments with hidden object detection	2019		Alexei A. Morozov and Olga S. Sushkova

Handwriting and character recognition

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Artificial Characters Dataset	Artificially generated data describing the structure of 10 capital English letters.	Coordinates of lines drawn given as integers. Various other features.	6000	Text	Handwriting recognition, classification	1992		H. Guvenir et al.
Letter Dataset	Upper case printed letters.	17 features are extracted from all images.	20,000	Text	OCR, classification	1991		D. Slate et al.
CASIA-HWDB	Offline handwritten Chinese character database. 3755 classes in the GB 2312 character set.	Gray-scaled images with background pixels labeled as 255.	1,172,907	Images, Text	Handwriting recognition, classification	2009		CASIA
CASIA-OLHWDB	Online handwritten Chinese character database, collected using Anoto pen on paper. 3755 classes in the GB 2312 character set.	Provides the sequences of coordinates of strokes.	1,174,364	Images, Text	Handwriting recognition, classification	2009		CASIA
Character Trajectories Dataset	Labeled samples of pen tip trajectories for people writing simple characters.	3-dimensional pen tip velocity trajectory matrix for each sample	2858	Text	Handwriting recognition, classification	2008		B. Williams
Chars74K Dataset	Character recognition in natural images of symbols used in both English and Kannada		74,107		Character recognition, handwriting recognition, OCR, classification	2009		T. de Campos
UJI Pen Characters Dataset	Isolated handwritten characters	Coordinates of pen position as characters were written given.	11,640	Text	Handwriting recognition, classification	2009		F. Prat et al.
Gisette Dataset	Handwriting samples from the often-confused 4 and 9 characters.	Features extracted from images, split into train/test, handwriting images size-normalized.	13,500	Images, text	Handwriting recognition, classification	2003		Yann LeCun et al.
Omniglot dataset	1623 different handwritten characters from 50 different alphabets.	Hand-labeled.	38,300	Images, text, strokes	Classification, one-shot learning	2015		American Association for the Advancement of Science
MNIST database	Database of handwritten digits.	Hand-labeled.	60,000	Images, text	Classification	1998		National Institute of Standards and Technology
Optical Recognition of Handwritten Digits Dataset	Normalized bitmaps of handwritten data.	Size normalized and mapped to bitmaps.	5620	Images, text	Handwriting recognition, classification	1998		E. Alpaydin et al.
Pen-Based Recognition of Handwritten Digits Dataset	Handwritten digits on electronic pen-tablet.	Feature vectors extracted to be uniformly spaced.	10,992	Images, text	Handwriting recognition, classification	1998		E. Alpaydin et al.
Semeion Handwritten Digit Dataset	Handwritten digits from 80 people.	All handwritten digits have been normalized for size and mapped to the same grid.	1593	Images, text	Handwriting recognition, classification	2008		T. Srl
HASYv2	Handwritten mathematical symbols	All symbols are centered and of size 32px x 32px.	168233	Images, text	Classification	2017		Martin Thoma
Noisy Handwritten Bangla Dataset	Includes Handwritten Numeral Dataset (10 classes) and Basic Character Dataset (50 classes), each dataset has three types of noise: white gaussian, motion blur, and reduced contrast.	All images are centered and of size 32x32.	Numeral Dataset: 23330, Character Dataset: 76000	Images, text	Handwriting recognition, classification	2017		M. Karki et al.

Aerial images

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Aerial Image Segmentation Dataset	80 high-resolution aerial images with spatial resolution ranging from 0.3 to 1.0.	Images manually segmented.	80	Images	Aerial Classification, object detection	2013		J. Yuan et al.
KIT AIS Data Set	Multiple labeled training and evaluation datasets of aerial images of crowds.	Images manually labeled to show paths of individuals through crowds.	~ 150	Images with paths	People tracking, aerial tracking	2012		M. Butenuth et al.
Wilt Dataset	Remote sensing data of diseased trees and other land cover.	Various features extracted.	4899	Images	Classification, aerial object detection	2014		B. Johnson
MASATI dataset	Maritime scenes of optical aerial images from the visible spectrum. It contains color images in dynamic marine environments, each image may contain one or multiple targets in different weather and illumination conditions.	Object bounding boxes and labeling.	7389	Images	Classification, aerial object detection	2018		A.-J. Gallego et al.
Forest Type Mapping Dataset	Satellite imagery of forests in Japan.	Image wavelength bands extracted.	326	Text	Classification	2015		B. Johnson
Overhead Imagery Research Data Set	Annotated overhead imagery. Images with multiple objects.	Over 30 annotations and over 60 statistics that describe the target within the context of the image.	1000	Images, text	Classification	2009		F. Tanner et al.
SpaceNet	SpaceNet is a corpus of commercial satellite imagery and labeled training data.	GeoTiff and GeoJSON files containing building footprints.	>17533	Images	Classification, Object Identification	2017		DigitalGlobe, Inc.
UC Merced Land Use Dataset	These images were manually extracted from large images from the USGS National Map Urban Area Imagery collection for various urban areas around the US.	This is a 21 class land use image dataset meant for research purposes. There are 100 images for each class.	2,100	Image chips of 256x256, 30 cm (1 foot) GSD	Land cover classification	2010		Yi Yang and Shawn Newsam
SAT-4 Airborne Dataset	Images were extracted from the National Agriculture Imagery Program (NAIP) dataset.	SAT-4 has four broad land cover classes, includes barren land, trees, grassland and a class that consists of all land cover classes other than the above three.	500,000	Images	Classification	2015		S. Basu et al.
SAT-6 Airborne Dataset	Images were extracted from the National Agriculture Imagery Program (NAIP) dataset.	SAT-6 has six broad land cover classes, includes barren land, trees, grassland, roads, buildings and water bodies.	405,000	Images	Classification	2015		S. Basu et al.

Other images

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Density functional theory quantum simulations of graphene	Labelled images of raw input to a simulation of graphene	Raw data (in HDF5 format) and output labels from density functional theory quantum simulation	60744 test and 501473 and training files	Labeled images	Regression	2019		K. Mills & I. Tamblyn
Quantum simulations of an electron in a two dimensional potential well	Labelled images of raw input to a simulation of 2d Quantum mechanics	Raw data (in HDF5 format) and output labels from quantum simulation	1.3 million images	Labeled images	Regression	2017		K. Mills, M.A. Spanner, & I. Tamblyn
MPII Cooking Activities Dataset	Videos and images of various cooking activities.	Activity paths and directions, labels, fine-grained motion labeling, activity class, still image extraction and labeling.	881,755 frames	Labeled video, images, text	Classification	2012		M. Rohrbach et al.
FAMOS Dataset	5,000 unique microstructures, all samples have been acquired 3 times with two different cameras.	Original PNG files, sorted per camera and then per acquisition. MATLAB datafiles with one 16384 times 5000 matrix per camera per acquisition.	30,000	Images and .mat files	Authentication	2012		S. Voloshynovskiy, et al.
PharmaPack Dataset	1,000 unique classes with 54 images per class.	Class labeling, many local descriptors, like SIFT and aKaZE, and local feature agreators, like Fisher Vector (FV).	54,000	Images and .mat files	Fine-grain classification	2017		O. Taran and S. Rezaeifar, et al.
Stanford Dogs Dataset	Images of 120 breeds of dogs from around the world.	Train/test splits and ImageNet annotations provided.	20,580	Images, text	Fine-grain classification	2011		A. Khosla et al.
StanfordExtra Dataset	2D keypoints and segmentations for the Stanford Dogs Dataset.	2D keypoints and segmentations provided.	12,035	Labelled images	3D reconstruction/pose estimation	2020		B. Biggs et al.
The Oxford-IIIT Pet Dataset	37 categories of pets with roughly 200 images of each.	Breed labeled, tight bounding box, foreground-background segmentation.	~ 7,400	Images, text	Classification, object detection	2012		O. Parkhi et al.
Corel Image Features Data Set	Database of images with features extracted.	Many features including color histogram, co-occurrence texture, and colormoments,	68,040	Text	Classification, object detection	1999		M. Ortega-Bindenberger et al.
Online Video Characteristics and Transcoding Time Dataset.	Transcoding times for various different videos and video properties.	Video features given.	168,286	Text	Regression	2015		T. Deneke et al.
Microsoft Sequential Image Narrative Dataset (SIND)	Dataset for sequential vision-to-language	Descriptive caption and storytelling given for each photo, and photos are arranged in sequences	81,743	Images, text	Visual storytelling	2016		Microsoft Research
Caltech-UCSD Birds-200-2011 Dataset	Large dataset of images of birds.	Part locations for birds, bounding boxes, 312 binary attributes given	11,788	Images, text	Classification	2011		C. Wah et al.
YouTube-8M	Large and diverse labeled video dataset	YouTube video IDs and associated labels from a diverse vocabulary of 4800 visual entities	8 million	Video, text	Video classification	2016		S. Abu-El-Haija et al.
YFCC100M	Large and diverse labeled image and video dataset	Flickr Videos and Images and associated description, titles, tags, and other metadata (such as EXIF and geotags)	100 million	Video, Image, Text	Video and Image classification	2016		B. Thomee et al.
Discrete LIRIS-ACCEDE	Short videos annotated for valence and arousal.	Valence and arousal labels.	9800	Video	Video emotion elicitation detection	2015		Y. Baveye et al.
Continuous LIRIS-ACCEDE	Long videos annotated for valence and arousal while also collecting Galvanic Skin Response.	Valence and arousal labels.	30	Video	Video emotion elicitation detection	2015		Y. Baveye et al.
MediaEval LIRIS-ACCEDE	Extension of Discrete LIRIS-ACCEDE including annotations for violence levels of the films.	Violence, valence and arousal labels.	10900	Video	Video emotion elicitation detection	2015		Y. Baveye et al.
Leeds Sports Pose	Articulated human pose annotations in 2000 natural sports images from Flickr.	Rough crop around single person of interest with 14 joint labels	2000	Images plus .mat file labels	Human pose estimation	2010		S. Johnson and M. Everingham
Leeds Sports Pose Extended Training	Articulated human pose annotations in 10,000 natural sports images from Flickr.	14 joint labels via crowdsourcing	10000	Images plus .mat file labels	Human pose estimation	2011		S. Johnson and M. Everingham
MCQ Dataset	6 different real multiple choice-based exams (735 answer sheets and 33,540 answer boxes) to evaluate computer vision techniques and systems developed for multiple choice test assessment systems.	None	735 answer sheets and 33,540 answer boxes	Images and .mat file labels	Development of multiple choice test assessment systems	2017		Afifi, M. et al.
Surveillance Videos	Real surveillance videos cover a large surveillance time (7 days with 24 hours each).	None	19 surveillance videos (7 days with 24 hours each).	Videos	Data compression	2016		Taj-Eddin, I. A. T. F. et al.
LILA BC	Labeled Information Library of Alexandria: Biology and Conservation. Labeled images that support machine learning research around ecology and environmental science.	None	~10M images	Images	Classification	2019		LILA working group
Can We See Photosynthesis?	32 videos for eight live and eight dead leaves recorded under both DC and AC lighting conditions.	None	32 videos	Videos	Liveness detection of plants	2017		Taj-Eddin, I. A. T. F. et al.

Text data

Datasets consisting primarily of text for tasks such as natural language processing, sentiment analysis, translation, and cluster analysis.

Reviews

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Amazon reviews	US product reviews from Amazon.com.	None.	~ 82M	Text	Classification, sentiment analysis	2015		McAuley et al.
OpinRank Review Dataset	Reviews of cars and hotels from Edmunds.com and TripAdvisor respectively.	None.	42,230 / ~259,000 respectively	Text	Sentiment analysis, clustering	2011		K. Ganesan et al.
MovieLens	22,000,000 ratings and 580,000 tags applied to 33,000 movies by 240,000 users.	None.	~ 22M	Text	Regression, clustering, classification	2016		GroupLens Research
Yahoo! Music User Ratings of Musical Artists	Over 10M ratings of artists by Yahoo users.	None described.	~ 10M	Text	Clustering, regression	2004		Yahoo!
Car Evaluation Data Set	Car properties and their overall acceptability.	Six categorical features given.	1728	Text	Classification	1997		M. Bohanec
YouTube Comedy Slam Preference Dataset	User vote data for pairs of videos shown on YouTube. Users voted on funnier videos.	Video metadata given.	1,138,562	Text	Classification	2012		Google
Skytrax User Reviews Dataset	User reviews of airlines, airports, seats, and lounges from Skytrax.	Ratings are fine-grain and include many aspects of airport experience.	41396	Text	Classification, regression	2015		Q. Nguyen
Teaching Assistant Evaluation Dataset	Teaching assistant reviews.	Features of each instance such as class, class size, and instructor are given.	151	Text	Classification	1997		W. Loh et al.
Vietnamese Students’ Feedback Corpus (UIT-VSFC)	Students’ Feedback.	Comments	16,000	Text	Classification	1997		Nguyen et al.
Vietnamese Social Media Emotion Corpus (UIT-VSMEC)	Users’ Facebook Comments.	Comments	6,927	Text	Classification	1997		Nguyen et al.

News articles

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
NYSK Dataset	English news articles about the case relating to allegations of sexual assault against the former IMF director Dominique Strauss-Kahn.	Filtered and presented in XML format.	10,421	XML, text	Sentiment analysis, topic extraction	2013		Dermouche, M. et al.
The Reuters Corpus Volume 1	Large corpus of Reuters news stories in English.	Fine-grain categorization and topic codes.	810,000	Text	Classification, clustering, summarization	2002		Reuters
The Reuters Corpus Volume 2	Large corpus of Reuters news stories in multiple languages.	Fine-grain categorization and topic codes.	487,000	Text	Classification, clustering, summarization	2005		Reuters
Thomson Reuters Text Research Collection	Large corpus of news stories.	Details not described.	1,800,370	Text	Classification, clustering, summarization	2009		T. Rose et al.
Saudi Newspapers Corpus	31,030 Arabic newspaper articles.	Metadata extracted.	31,030	JSON	Summarization, clustering	2015		M. Alhagri
RE3D (Relationship and Entity Extraction Evaluation Dataset)	Entity and Relation marked data from various news and government sources. Sponsored by Dstl	Filtered, categorisation using Baleen types	not known	JSON	Classification, Entity and Relation recognition	2017		Dstl
Examiner Spam Clickbait Catalogue	Clickbait, spam, crowd-sourced headlines from 2010 to 2015	Publish date and headlines	3,089,781	CSV	Clustering, Events, Sentiment	2016		R. Kulkarni
ABC Australia News Corpus	Entire news corpus of ABC Australia from 2003 to 2019	Publish date and headlines	1,186,018	CSV	Clustering, Events, Sentiment	2020		R. Kulkarni
Worldwide News - Aggregate of 20K Feeds	One week snapshot of all online headlines in 20+ languages	Publish time, URL and headlines	1,398,431	CSV	Clustering, Events, Language Detection	2018		R. Kulkarni
Reuters News Wire Headline	11 Years of timestamped events published on the news-wire	Publish time, Headline Text	16,121,310	CSV	NLP, Computational Linguistics, Events	2018		R. Kulkarni
The Irish Times Ireland News Corpus	24 Years of Ireland News from 1996 to 2019	Publish time, Headline Category and Text	1,484,340	CSV	NLP, Computational Linguistics, Events	2020		R. Kulkarni
News Headlines Dataset for Sarcasm Detection	High quality dataset with Sarcastic and Non-sarcastic news headlines.	Clean, normalized text	26,709	JSON	NLP, Classification, Linguistics	2018		Rishabh Misra

Messages

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Enron Email Dataset	Emails from employees at Enron organized into folders.	Attachments removed, invalid email addresses converted to user@enron.com or no_address@enron.com.	~ 500,000	Text	Network analysis, sentiment analysis	2004 (2015)		Klimt, B. and Y. Yang
Ling-Spam Dataset	Corpus containing both legitimate and spam emails.	Four version of the corpus involving whether or not a lemmatiser or stop-list was enabled.	2,412 Ham 481 Spam	Text	Classification	2000		Androutsopoulos, J. et al.
SMS Spam Collection Dataset	Collected SMS spam messages.	None.	5,574	Text	Classification	2011		T. Almeida et al.
Twenty Newsgroups Dataset	Messages from 20 different newsgroups.	None.	20,000	Text	Natural language processing	1999		T. Mitchell et al.
Spambase Dataset	Spam emails.	Many text features extracted.	4,601	Text	Spam detection, classification	1999		M. Hopkins et al.
ColBERT Dataset	Short jokes.	Outliers removed.	200,000	Text	Humor detection, classification	2020		I. Annamoradnejad.

Twitter and tweets

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
MovieTweetings	Movie rating dataset based on public and well-structured tweets		~710,000	Text	Classification, regression	2018		S. Dooms
Twitter100k	Pairs of images and tweets		100,000	Text and Images	Cross-media retrieval	2017		Y. Hu, et al.
Sentiment140	Tweet data from 2009 including original text, time stamp, user and sentiment.	Classified using distant supervision from presence of emoticon in tweet.	1,578,627	Tweets, comma, separated values	Sentiment analysis	2009		A. Go et al.
ASU Twitter Dataset	Twitter network data, not actual tweets. Shows connections between a large number of users.	None.	11,316,811 users, 85,331,846 connections	Text	Clustering, graph analysis	2009		R. Zafarani et al.
SNAP Social Circles: Twitter Database	Large Twitter network data.	Node features, circles, and ego networks.	1,768,149	Text	Clustering, graph analysis	2012		J. McAuley et al.
Twitter Dataset for Arabic Sentiment Analysis	Arabic tweets.	Samples hand-labeled as positive or negative.	2000	Text	Classification	2014		N. Abdulla
Buzz in Social Media Dataset	Data from Twitter and Tom's Hardware. This dataset focuses on specific buzz topics being discussed on those sites.	Data is windowed so that the user can attempt to predict the events leading up to social media buzz.	140,000	Text	Regression, Classification	2013		F. Kawala et al.
Paraphrase and Semantic Similarity in Twitter (PIT)	This dataset focuses on whether tweets have (almost) same meaning/information or not. Manually labeled.	tokenization, part-of-speech and named entity tagging	18,762	Text	Regression, Classification	2015		Xu et al.
Geoparse Twitter benchmark dataset	This dataset contains tweets during different news events in different countries. Manually labeled location mentions.	location annotations added to JSON metadata	6,386	Tweets, JSON	Classification, Information Extraction	2014		S.E. Middleton et al.
Dutch Social media collection	This dataset contains Covid-19 tweets made by Dutch speakers or users from Netherlands. The data has been machine-annotated	classified for sentiment, tweet text & user description translated to English. Industry mention are extracted	271,342	JSONL	Sentiment, multi-label classification, machine translation	2020		Aaaksh Gupta, CoronaWhy

Dialogues

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
NPS Chat Corpus	Posts from age-specific online chat rooms.	Hand privacy masked, tagged for part of speech and dialogue-act.	~ 500,000	XML	NLP, programming, linguistics	2007		Forsyth, E., Lin, J., & Martell, C.
Twitter Triple Corpus	A-B-A triples extracted from Twitter.		4,232	Text	NLP	2016		Sordini, A. et al.
UseNet Corpus	UseNet forum postings.	Anonymized e-mails and URLs. Omitted documents with lengths <500 words or >500,000 words, or that were <90% English.	7 billion	Text		2011		Shaoul, C., & Westbury C.
NUS SMS Corpus	SMS messages collected between two users, with timing analysis.		~ 10,000	XML	NLP	2011		KAN, M
Reddit All Comments Corpus	All Reddit comments (as of 2015).		~ 1.7 billion	JSON	NLP, research	2015		Stuck_In_the_Matrix
Ubuntu Dialogue Corpus	Dialogues extracted from Ubuntu chat stream on IRC.			CSV	Dialogue Systems Research	2015		Lowe, R. et al.

Other text

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Web of Science Dataset	Hierarchical Datasets for Text Classification	None.	46,985	Text	Classification, Categorization	2017		K. Kowsari et al.
Legal Case Reports	Federal Court of Australia cases from 2006 to 2009.	None.	4,000	Text	Summarization, citation analysis	2012		F. Galgani et al.
Blogger Authorship Corpus	Blog entries of 19,320 people from blogger.com.	Blogger self-provided gender, age, industry, and astrological sign.	681,288	Text	Sentiment analysis, summarization, classification	2006		J. Schler et al.
Social Structure of Facebook Networks	Large dataset of the social structure of Facebook.	None.	100 colleges covered	Text	Network analysis, clustering	2012		A. Traud et al.
Dataset for the Machine Comprehension of Text	Stories and associated questions for testing comprehension of text.	None.	660	Text	Natural language processing, machine comprehension	2013		M. Richardson et al.
The Penn Treebank Project	Naturally occurring text annotated for linguistic structure.	Text is parsed into semantic trees.	~ 1M words	Text	Natural language processing, summarization	1995		M. Marcus et al.
DEXTER Dataset	Task given is to determine, from features given, which articles are about corporate acquisitions.	Features extracted include word stems. Distractor features included.	2600	Text	Classification	2008		Reuters
Google Books N-grams	N-grams from a very large corpus of books	None.	2.2 TB of text	Text	Classification, clustering, regression	2011		Google
Personae Corpus	Collected for experiments in Authorship Attribution and Personality Prediction. Consists of 145 Dutch-language essays.	In addition to normal texts, syntactically annotated texts are given.	145	Text	Classification, regression	2008		K. Luyckx et al.
CNAE-9 Dataset	Categorization task for free text descriptions of Brazilian companies.	Word frequency has been extracted.	1080	Text	Classification	2012		P. Ciarelli et al.
Sentiment Labeled Sentences Dataset	3000 sentiment labeled sentences.	Sentiment of each sentence has been hand labeled as positive or negative.	3000	Text	Classification, sentiment analysis	2015		D. Kotzias
BlogFeedback Dataset	Dataset to predict the number of comments a post will receive based on features of that post.	Many features of each post extracted.	60,021	Text	Regression	2014		K. Buza
Stanford Natural Language Inference (SNLI) Corpus	Image captions matched with newly constructed sentences to form entailment, contradiction, or neutral pairs.	Entailment class labels, syntactic parsing by the Stanford PCFG parser	570,000	Text	Natural language inference/recognizing textual entailment	2015		S. Bowman et al.
DSL Corpus Collection (DSLCC)	A multilingual collection of short excerpts of journalistic texts in similar languages and dialects.	None	294,000 phrases	Text	Discriminating between similar languages	2017		Tan, Liling et al.
Urban Dictionary Dataset	Corpus of words, votes and definitions	User names anonymised	2,580,925	CSV	NLP, Machine comprehension	2016 May		Anonymous
T-REx	Wikipedia abstracts aligned with Wikidata entities	Alignment of Wikidata triples with Wikipedia abstracts	11M aligned triples	JSON and NIF [1]	NLP, Relation Extraction	2018		H. Elsahar et al.
General Language Understanding Evaluation (GLUE)	Benchmark of nine tasks	Various	~1M sentences and sentence pairs		NLU	2018		Wang et al.
Atticus Open Contract Dataset (AOK)	Dataset of legal contracts with rich expert annotations		~3,000 labels	CSV and PDF	Natural language processing, QnA	2020		The Atticus Project
Vietnamese Image Captioning Dataset (UIT-ViIC)	Vietnamese Image Captioning Dataset		19,250 captions for 3,850 images	CSV and PDF	Natural language processing, Computer vision	2020		Lam et al.
Vietnamese Names annotated with Genders (UIT-ViNames)	Vietnamese Names annotated with Genders		26,850 Vietnamese full names annotated with genders	CSV	Natural language processing	2020		To et al.

Sound data

Datasets of sounds and sound features.

Speech

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Zero Resource Speech Challenge 2015	Spontaneous speech (English), Read speech (Xitsonga).	raw wav	English: 5h, 12 speakers; Xitsonga: 2h30; 24 speakers	sound	Unsupervised discovery of speech features/subword units/word units	2015		Versteegh et al.
Parkinson Speech Dataset	Multiple recordings of people with and without Parkinson's Disease.	Voice features extracted, disease scored by physician using unified Parkinson's disease rating scale	1,040	Text	Classification, regression	2013		B. E. Sakar et al.
Spoken Arabic Digits	Spoken Arabic digits from 44 male and 44 female.	Time-series of mel-frequency cepstrum coefficients.	8,800	Text	Classification	2010		M. Bedda et al.
ISOLET Dataset	Spoken letter names.	Features extracted from sounds.	7797	Text	Classification	1994		R. Cole et al.
Japanese Vowels Dataset	Nine male speakers uttered two Japanese vowels successively.	Applied 12-degree linear prediction analysis to it to obtain a discrete-time series with 12 cepstrum coefficients.	640	Text	Classification	1999		M. Kudo et al.
Parkinson's Telemonitoring Dataset	Multiple recordings of people with and without Parkinson's Disease.	Sound features extracted.	5875	Text	Classification	2009		A. Tsanas et al.
TIMIT	Recordings of 630 speakers of eight major dialects of American English, each reading ten phonetically rich sentences.	Speech is lexically and phonemically transcribed.	6300	Text	Speech recognition, classification.	1986		J. Garofolo et al.
Arabic Speech Corpus	A single-speaker, Modern Standard Arabic (MSA) speech corpus with phonetic and orthographic transcripts aligned to phoneme level	Speech is orthographically and phonetically transcribed with stress marks.	~1900	Text, WAV	Speech Synthesis, Speech Recognition, Corpus Alignment, Speech Therapy, Education.	2016		N. Halabi
Common Voice	A public domain database of crowdsourced data across a wide range of dialects.	Validation by other users	English: 1,118 hours	MP3 with corresponding text files	Speech recognition	June 2017 (December 2019)		Mozilla

Music

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Geographic Origin of Music Data Set	Audio features of music samples from different locations.	Audio features extracted using MARSYAS software.	1,059	Text	Geographic classification, clustering	2014		F. Zhou et al.
Million Song Dataset	Audio features from one million different songs.	Audio features extracted.	1M	Text	Classification, clustering	2011		T. Bertin-Mahieux et al.
MUSDB18	Multi-track popular music recordings	Raw audio	150	MP4, WAV	Source Separation	2017		Z. Rafii et al.
Free Music Archive	Audio under Creative Commons from 100k songs (343 days, 1TiB) with a hierarchy of 161 genres, metadata, user data, free-form text.	Raw audio and audio features.	106,574	Text, MP3	Classification, recommendation	2017		M. Defferrard et al.
Bach Choral Harmony Dataset	Bach chorale chords.	Audio features extracted.	5665	Text	Classification	2014		D. Radicioni et al.

Other sounds

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
UrbanSound	Labeled sound recordings of sounds like air conditioners, car horns and children playing.	Sorted into folders by class of events as well as metadata in a JSON file and annotations in a CSV file.	1,059	Sound (WAV)	Classification	2014		J. Salamon et al.
AudioSet	10-second sound snippets from YouTube videos, and an ontology of over 500 labels.	128-d PCA'd VGG-ish features every 1 second.	2,084,320	Text (CSV) and TensorFlow Record files	Classification	2017		J. Gemmeke et al., Google
Bird Audio Detection challenge	Audio from environmental monitoring stations, plus crowdsourced recordings		17,000+		Classification	2016 (2018)		Queen Mary University and IEEE Signal Processing Society
WSJ0 Hipster Ambient Mixtures	Audio from WSJ0 mixed with noise recorded in the San Francisco Bay Area	Noise clips matched to WSJ0 clips	28,000	Sound (WAV)	Audio source separation	2019		Wichern, G., et al., Whisper and MERL
Clotho	4,981 audio samples of 15 to 30 seconds long, each audio sample having five different captions of eight to 20 words long.		24,905	Sound (WAV) and text (CSV)	Automated audio captioning	2020		K. Drossos, S. Lipping, and T. Virtanen

Signal data

Datasets containing electric signal information requiring some sort of Signal processing for further analysis.

Electrical

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Witty Worm Dataset	Dataset detailing the spread of the Witty worm and the infected computers.	Split into a publicly available set and a restricted set containing more sensitive information like IP and UDP headers.	55,909 IP addresses	Text	Classification	2004		Center for Applied Internet Data Analysis
Cuff-Less Blood Pressure Estimation Dataset	Cleaned vital signals from human patients which can be used to estimate blood pressure.	125 Hz vital signs have been cleaned.	12,000	Text	Classification, regression	2015		M. Kachuee et al.
Gas Sensor Array Drift Dataset	Measurements from 16 chemical sensors utilized in simulations for drift compensation.	Extensive number of features given.	13,910	Text	Classification	2012		A. Vergara
Servo Dataset	Data covering the nonlinear relationships observed in a servo-amplifier circuit.	Levels of various components as a function of other components are given.	167	Text	Regression	1993		K. Ullrich
UJIIndoorLoc-Mag Dataset	Indoor localization database to test indoor positioning systems. Data is magnetic field based.	Train and test splits given.	40,000	Text	Classification, regression, clustering	2015		D. Rambla et al.
Sensorless Drive Diagnosis Dataset	Electrical signals from motors with defective components.	Statistical features extracted.	58,508	Text	Classification	2015		M. Bator

Motion-tracking

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Wearable Computing: Classification of Body Postures and Movements (PUC-Rio)	People performing five standard actions while wearing motion trackers.	None.	165,632	Text	Classification	2013		Pontifical Catholic University of Rio de Janeiro
Gesture Phase Segmentation Dataset	Features extracted from video of people doing various gestures.	Features extracted aim at studying gesture phase segmentation.	9900	Text	Classification, clustering	2014		R. Madeo et a
Vicon Physical Action Data Set Dataset	10 normal and 10 aggressive physical actions that measure the human activity tracked by a 3D tracker.	Many parameters recorded by 3D tracker.	3000	Text	Classification	2011		T. Theodoridis
Daily and Sports Activities Dataset	Motor sensor data for 19 daily and sports activities.	Many sensors given, no preprocessing done on signals.	9120	Text	Classification	2013		B. Barshan et al.
Human Activity Recognition Using Smartphones Dataset	Gyroscope and accelerometer data from people wearing smartphones and performing normal actions.	Actions performed are labeled, all signals preprocessed for noise.	10,299	Text	Classification	2012		J. Reyes-Ortiz et al.
Australian Sign Language Signs	Australian sign language signs captured by motion-tracking gloves.	None.	2565	Text	Classification	2002		M. Kadous
Weight Lifting Exercises monitored with Inertial Measurement Units	Five variations of the biceps curl exercise monitored with IMUs.	Some statistics calculated from raw data.	39,242	Text	Classification	2013		W. Ugulino et al.
sEMG for Basic Hand movements Dataset	Two databases of surface electromyographic signals of 6 hand movements.	None.	3000	Text	Classification	2014		C. Sapsanis et al.
REALDISP Activity Recognition Dataset	Evaluate techniques dealing with the effects of sensor displacement in wearable activity recognition.	None.	1419	Text	Classification	2014		O. Banos et al.
Heterogeneity Activity Recognition Dataset	Data from multiple different smart devices for humans performing various activities.	None.	43,930,257	Text	Classification, clustering	2015		A. Stisen et al.
Indoor User Movement Prediction from RSS Data	Temporal wireless network data that can be used to track the movement of people in an office.	None.	13,197	Text	Classification	2016		D. Bacciu
PAMAP2 Physical Activity Monitoring Dataset	18 different types of physical activities performed by 9 subjects wearing 3 IMUs.	None.	3,850,505	Text	Classification	2012		A. Reiss
OPPORTUNITY Activity Recognition Dataset	Human Activity Recognition from wearable, object, and ambient sensors is a dataset devised to benchmark human activity recognition algorithms.	None.	2551	Text	Classification	2012		D. Roggen et al.
Real World Activity Recognition Dataset	Human Activity Recognition from wearable devices. Distinguishes between seven on-body device positions and comprises six different kinds of sensors.	None.	3,150,000 (per sensor)	Text	Classification	2016		T. Sztyler et al.
Toronto Rehab Stroke Pose Dataset	3D human pose estimates (Kinect) of stroke patients and healthy participants performing a set of tasks using a stroke rehabilitation robot.	None.	10 healthy person and 9 stroke survivors (3500-6000 frames per person)	CSV	Classification	2017		E. Dolatabadi et al.
Corpus of Social Touch (CoST)	7805 gesture captures of 14 different social touch gestures performed by 31 subjects. The gestures were performed in three variations: gentle, normal and rough, on a pressure sensor grid wrapped around a mannequin arm.	Touch gestures performed are segmented and labeled.	7805 gesture captures	CSV	Classification	2016		M. Jung et al.

Other signals

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Wine Dataset	Chemical analysis of wines grown in the same region in Italy but derived from three different cultivars.	13 properties of each wine are given	178	Text	Classification, regression	1991		M. Forina et al.
Combined Cycle Power Plant Data Set	Data from various sensors within a power plant running for 6 years.	None	9568	Text	Regression	2014		P. Tufekci et al.

Physical data

Datasets from physical systems.

High-energy physics

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
HIGGS Dataset	Monte Carlo simulations of particle accelerator collisions.	28 features of each collision are given.	11M	Text	Classification	2014		D. Whiteson
HEPMASS Dataset	Monte Carlo simulations of particle accelerator collisions. Goal is to separate the signal from noise.	28 features of each collision are given.	10,500,000	Text	Classification	2016		D. Whiteson

Systems

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Yacht Hydrodynamics Dataset	Yacht performance based on dimensions.	Six features are given for each yacht.	308	Text	Regression	2013		R. Lopez
Robot Execution Failures Dataset	5 data sets that center around robotic failure to execute common tasks.	Integer valued features such as torque and other sensor measurements.	463	Text	Classification	1999		L. Seabra et al.
Pittsburgh Bridges Dataset	Design description is given in terms of several properties of various bridges.	Various bridge features are given.	108	Text	Classification	1990		Y. Reich et al.
Automobile Dataset	Data about automobiles, their insurance risk, and their normalized losses.	Car features extracted.	205	Text	Regression	1987		J. Schimmer et al.
Auto MPG Dataset	MPG data for cars.	Eight features of each car given.	398	Text	Regression	1993		Carnegie Mellon University
Energy Efficiency Dataset	Heating and cooling requirements given as a function of building parameters.	Building parameters given.	768	Text	Classification, regression	2012		A. Xifara et al.
Airfoil Self-Noise Dataset	A series of aerodynamic and acoustic tests of two and three-dimensional airfoil blade sections.	Data about frequency, angle of attack, etc., are given.	1503	Text	Regression	2014		R. Lopez
Challenger USA Space Shuttle O-Ring Dataset	Attempt to predict O-ring problems given past Challenger data.	Several features of each flight, such as launch temperature, are given.	23	Text	Regression	1993		D. Draper et al.
Statlog (Shuttle) Dataset	NASA space shuttle datasets.	Nine features given.	58,000	Text	Classification	2002		NASA

Astronomy

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Volcanoes on Venus – JARtool experiment Dataset	Venus images returned by the Magellan spacecraft.	Images are labeled by humans.	not given	Images	Classification	1991		M. Burl
MAGIC Gamma Telescope Dataset	Monte Carlo generated high-energy gamma particle events.	Numerous features extracted from the simulations.	19,020	Text	Classification	2007		R. Bock
Solar Flare Dataset	Measurements of the number of certain types of solar flare events occurring in a 24-hour period.	Many solar flare-specific features are given.	1389	Text	Regression, classification	1989		G. Bradshaw

Earth science

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Volcanoes of the World	Volcanic eruption data for all known volcanic events on earth.	Details such as region, subregion, tectonic setting, dominant rock type are given.	1535	Text	Regression, classification	2013		E. Venzke et al.
Seismic-bumps Dataset	Seismic activities from a coal mine.	Seismic activity was classified as hazardous or not.	2584	Text	Classification	2013		M. Sikora et al.

Other physical

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Concrete Compressive Strength Dataset	Dataset of concrete properties and compressive strength.	Nine features are given for each sample.	1030	Text	Regression	2007		I. Yeh
Concrete Slump Test Dataset	Concrete slump flow given in terms of properties.	Features of concrete given such as fly ash, water, etc.	103	Text	Regression	2009		I. Yeh
Musk Dataset	Predict if a molecule, given the features, will be a musk or a non-musk.	168 features given for each molecule.	6598	Text	Classification	1994		Arris Pharmaceutical Corp.
Steel Plates Faults Dataset	Steel plates of 7 different types.	27 features given for each sample.	1941	Text	Classification	2010		Semeion Research Center

Biological data

Datasets from biological systems.

Human

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
EEG Database	Study to examine EEG correlates of genetic predisposition to alcoholism.	Measurements from 64 electrodes placed on the scalp sampled at 256 Hz (3.9 ms epoch) for 1 second.	122	Text	Classification	1999		H. Begleiter
P300 Interface Dataset	Data from nine subjects collected using P300-based brain-computer interface for disabled subjects.	Split into four sessions for each subject. MATLAB code given.	1,224	Text	Classification	2008		U. Hoffman et al.
Heart Disease Data Set	Attributed of patients with and without heart disease.	75 attributes given for each patient with some missing values.	303	Text	Classification	1988		A. Janosi et al.
Breast Cancer Wisconsin (Diagnostic) Dataset	Dataset of features of breast masses. Diagnoses by physician is given.	10 features for each sample are given.	569	Text	Classification	1995		W. Wolberg et al.
National Survey on Drug Use and Health	Large scale survey on health and drug use in the United States.	None.	55,268	Text	Classification, regression	2012		United States Department of Health and Human Services
Lung Cancer Dataset	Lung cancer dataset without attribute definitions	56 features are given for each case	32	Text	Classification	1992		Z. Hong et al.
Arrhythmia Dataset	Data for a group of patients, of which some have cardiac arrhythmia.	276 features for each instance.	452	Text	Classification	1998		H. Altay et al.
Diabetes 130-US hospitals for years 1999–2008 Dataset	9 years of readmission data across 130 US hospitals for patients with diabetes.	Many features of each readmission are given.	100,000	Text	Classification, clustering	2014		J. Clore et al.
Diabetic Retinopathy Debrecen Dataset	Features extracted from images of eyes with and without diabetic retinopathy.	Features extracted and conditions diagnosed.	1151	Text	Classification	2014		B. Antal et al.
Diabetic Retinopathy Messidor Dataset	Methods to evaluate segmentation and indexing techniques in the field of retinal ophthalmology (MESSIDOR)	Features retinopathy grade and risk of macular edema	1200	Images, Text	Classification, Segmentation	2008		Messidor Project
Liver Disorders Dataset	Data for people with liver disorders.	Seven biological features given for each patient.	345	Text	Classification	1990		Bupa Medical Research Ltd.
Thyroid Disease Dataset	10 databases of thyroid disease patient data.	None.	7200	Text	Classification	1987		R. Quinlan
Mesothelioma Dataset	Mesothelioma patient data.	Large number of features, including asbestos exposure, are given.	324	Text	Classification	2016		A. Tanrikulu et al.
Parkinson's Vision-Based Pose Estimation Dataset	2D human pose estimates of Parkinson's patients performing a variety of tasks.	Camera shake has been removed from trajectories.	134	Text	Classification, regression	2017		M. Li et al.
KEGG Metabolic Reaction Network (Undirected) Dataset	Network of metabolic pathways. A reaction network and a relation network are given.	Detailed features for each network node and pathway are given.	65,554	Text	Classification, clustering, regression	2011		M. Naeem et al.
Modified Human Sperm Morphology Analysis Dataset (MHSMA)	Human sperm images from 235 patients with male factor infertility, labeled for normal or abnormal sperm acrosome, head, vacuole, and tail.	Cropped around single sperm head. Magnification normalized. Training, validation, and test set splits created.	1,540	.npy files	Classification	2019		S. Javadi and S.A. Mirroshandel

Animal

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Abalone Dataset	Physical measurements of Abalone. Weather patterns and location are also given.	None.	4177	Text	Regression	1995		Marine Research Laboratories – Taroona
Zoo Dataset	Artificial dataset covering 7 classes of animals.	Animals are classed into 7 categories and features are given for each.	101	Text	Classification	1990		R. Forsyth
Demospongiae Dataset	Data about marine sponges.	503 sponges in the Demosponge class are described by various features.	503	Text	Classification	2010		E. Armengol et al.
Splice-junction Gene Sequences Dataset	Primate splice-junction gene sequences (DNA) with associated imperfect domain theory.	None.	3190	Text	Classification	1992		G. Towell et al.
Mice Protein Expression Dataset	Expression levels of 77 proteins measured in the cerebral cortex of mice.	None.	1080	Text	Classification, Clustering	2015		C. Higuera et al.

Plant

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Forest Fires Dataset	Forest fires and their properties.	13 features of each fire are extracted.	517	Text	Regression	2008		P. Cortez et al.
Iris Dataset	Three types of iris plants are described by 4 different attributes.	None.	150	Text	Classification	1936		R. Fisher
Plant Species Leaves Dataset	Sixteen samples of leaf each of one-hundred plant species.	Shape descriptor, fine-scale margin, and texture histograms are given.	1600	Text	Classification	2012		J. Cope et al.
Mushroom Dataset	Mushroom attributes and classification.	Many properties of each mushroom are given.	8124	Text	Classification	1987		J. Schlimmer
Soybean Dataset	Database of diseased soybean plants.	35 features for each plant are given. Plants are classified into 19 categories.	307	Text	Classification	1988		R. Michalski et al.
Seeds Dataset	Measurements of geometrical properties of kernels belonging to three different varieties of wheat.	None.	210	Text	Classification, clustering	2012		Charytanowicz et al.
Covertype Dataset	Data for predicting forest cover type strictly from cartographic variables.	Many geographical features given.	581,012	Text	Classification	1998		J. Blackard et al.
Abscisic Acid Signaling Network Dataset	Data for a plant signaling network. Goal is to determine set of rules that governs the network.	None.	300	Text	Causal-discovery	2008		J. Jenkens et al.
Folio Dataset	20 photos of leaves for each of 32 species.	None.	637	Images, text	Classification, clustering	2015		T. Munisami et al.
Oxford Flower Dataset	17 category dataset of flowers.	Train/test splits, labeled images,	1360	Images, text	Classification	2006		M-E Nilsback et al.
Plant Seedlings Dataset	12 category dataset of plant seedlings.	Labelled images, segmented images,	5544	Images	Classification, detection	2017		Giselsson et al.
Fruits 360 dataset	Database with images of 120 fruits and vegetables.	100x100 pixels, White background.	82213	Images (jpg)	Classification	2017-2019		Mihai Oltean, Horea Muresan

Microbe

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Ecoli Dataset	Protein localization sites.	Various features of the protein localizations sites are given.	336	Text	Classification	1996		K. Nakai et al.
MicroMass Dataset	Identification of microorganisms from mass-spectrometry data.	Various mass spectrometer features.	931	Text	Classification	2013		P. Mahe et al.
Yeast Dataset	Predictions of Cellular localization sites of proteins.	Eight features given per instance.	1484	Text	Classification	1996		K. Nakai et al.

Drug Discovery

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Tox21 Dataset	Prediction of outcome of biological assays.	Chemical descriptors of molecules are given.	12707	Text	Classification	2016		A. Mayr et al.

Anomaly data

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Numenta Anomaly Benchmark (NAB)	Data are ordered, timestamped, single-valued metrics. All data files contain anomalies, unless otherwise noted.	None	50+ files	Comma separated values	Anomaly detection	2016 (continually updated)		Numenta
Skoltech Anomaly Benchmark (SKAB)	Each file represents a single experiment and contains a single anomaly. The dataset represents a multivariate time series collected from the sensors installed on the testbed.	There are two markups for Outlier detection (point anomalies) and Changepoint detection (collective anomalies) problems	30+ files (v0.9)	Comma separated values	Anomaly detection	2020 (continually updated)		Iurii D. Katser and Vyacheslav O. Kozitsin
On the Evaluation of Unsupervised Outlier Detection: Measures, Datasets, and an Empirical Study	Most data files are adapted from UCI Machine Learning Repository data, some are collected from the literature.	treated for missing values, numerical attributes only, different percentages of anomalies, labels	1000+ files	ARFF	Anomaly detection	2016 (possibly updated with new datasets and/or results)		Campos et al.

Question Answering data

This section includes datasets that deals with structured data.

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
DBpedia Neural Question Answering (DBNQA) Dataset	A large collection of Question to SPARQL specially design for Open Domain Neural Question Answering over DBpedia Knowledgebase.	This dataset contains a large collection of Open Neural SPARQL Templates and instances for training Neural SPARQL Machines; it was pre-processed by semi-automatic annotation tools as well as by three SPARQL experts.	894,499	Question-query pairs	Question Answering	2018		Hartmann, Soru, and Marx et al.
Vietnamese Question Answering Dataset (UIT-ViQuAD)	A large collection of Vietnamese questions for evaluating MRC models.	This dataset comprises over 23,000 human-generated question-answer pairs based on 5,109 passages of 174 Vietnamese articles from Wikipedia.	23,074	Question-answer pairs	Question Answering	2020		Nguyen et al.
Vietnamese Multiple-Choice Machine Reading Comprehension Corpus(ViMMRC)	A collection of Vietnamese multiple-choice questions for evaluating MRC models.	This corpus includes 2,783 Vietnamese multiple-choice questions.	2,783	Question-answer pairs	Question Answering/Machine Reading Comprehension	2020		Nguyen et al.

Multivariate data

Datasets consisting of rows of observations and columns of attributes characterizing those observations. Typically used for regression analysis or classification but other types of algorithms can also be used. This section includes datasets that do not fit in the above categories.

Financial

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Dow Jones Index	Weekly data of stocks from the first and second quarters of 2011.	Calculated values included such as percentage change and a lags.	750	Comma separated values	Classification, regression, Time series	2014		M. Brown et al.
Statlog (Australian Credit Approval)	Credit card applications either accepted or rejected and attributes about the application.	Attribute names are removed as well as identifying information. Factors have been relabeled.	690	Comma separated values	Classification	1987		R. Quinlan
eBay auction data	Auction data from various eBay.com objects over various length auctions	Contains all bids, bidderID, bid times, and opening prices.	~ 550	Text	Regression, classification	2012		G. Shmueli et al.
Statlog (German Credit Data)	Binary credit classification into "good" or "bad" with many features	Various financial features of each person are given.	690	Text	Classification	1994		H. Hofmann
Bank Marketing Dataset	Data from a large marketing campaign carried out by a large bank .	Many attributes of the clients contacted are given. If the client subscribed to the bank is also given.	45,211	Text	Classification	2012		S. Moro et al.
Istanbul Stock Exchange Dataset	Several stock indexes tracked for almost two years.	None.	536	Text	Classification, regression	2013		O. Akbilgic
Default of Credit Card Clients	Credit default data for Taiwanese creditors.	Various features about each account are given.	30,000	Text	Classification	2016		I. Yeh

Weather

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Cloud DataSet	Data about 1024 different clouds.	Image features extracted.	1024	Text	Classification, clustering	1989		P. Collard
El Nino Dataset	Oceanographic and surface meteorological readings taken from a series of buoys positioned throughout the equatorial Pacific.	12 weather attributes are measured at each buoy.	178080	Text	Regression	1999		Pacific Marine Environmental Laboratory
Greenhouse Gas Observing Network Dataset	Time-series of greenhouse gas concentrations at 2921 grid cells in California created using simulations of the weather.	None.	2921	Text	Regression	2015		D. Lucas
Atmospheric CO2 from Continuous Air Samples at Mauna Loa Observatory	Continuous air samples in Hawaii, USA. 44 years of records.	None.	44 years	Text	Regression	2001		Mauna Loa Observatory
Ionosphere Dataset	Radar data from the ionosphere. Task is to classify into good and bad radar returns.	Many radar features given.	351	Text	Classification	1989		Johns Hopkins University
Ozone Level Detection Dataset	Two ground ozone level datasets.	Many features given, including weather conditions at time of measurement.	2536	Text	Classification	2008		K. Zhang et al.

Census

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Adult Dataset	Census data from 1994 containing demographic features of adults and their income.	Cleaned and anonymized.	48,842	Comma separated values	Classification	1996		United States Census Bureau
Census-Income (KDD)	Weighted census data from the 1994 and 1995 Current Population Surveys.	Split into training and test sets.	299,285	Comma separated values	Classification	2000		United States Census Bureau
IPUMS Census Database	Census data from the Los Angeles and Long Beach areas.	None	256,932	Text	Classification, regression	1999		IPUMS
US Census Data 1990	Partial data from 1990 US census.	Results randomized and useful attributes selected.	2,458,285	Text	Classification, regression	1990		United States Census Bureau

Transit

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Bike Sharing Dataset	Hourly and daily count of rental bikes in a large city.	Many features, including weather, length of trip, etc., are given.	17,389	Text	Regression	2013		H. Fanaee-T
New York City Taxi Trip Data	Trip data for yellow and green taxis in New York City.	Gives pick up and drop off locations, fares, and other details of trips.	6 years	Text	Classification, clustering	2015		New York City Taxi and Limousine Commission
Taxi Service Trajectory ECML PKDD	Trajectories of all taxis in a large city.	Many features given, including start and stop points.	1,710,671	Text	Clustering, causal-discovery	2015		M. Ferreira et al.
METR-LA	Speed from loop detectors in the highway of Los Angles County.	Average speed in 5 minutes timesteps.	7,094,304 from 207 sensors and 34,272 timesteps	Comma separated values	Regression, Forecasting	2014		Jagadish et. al.
PeMS	Speed, flow, occupancy and other metrics from loop detectors and other sensors in the freeway of the State of California, U.S.A..	Metric usually aggregated via Average into 5 minutes timesteps.	39,000 individual detectors, each containing years of timeseries	Comma separated values	Regression, Forecasting, Nowcasting, Interpolation	(updated realtime)		California Department of Transportation

Internet

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Webpages from Common Crawl 2012	Large collection of webpages and how they are connected via hyperlinks	None.	3.5B	Text	clustering, classification	2013		V. Granville
Internet Advertisements Dataset	Dataset for predicting if a given image is an advertisement or not.	Features encode geometry of ads and phrases occurring in the URL.	3279	Text	Classification	1998		N. Kushmerick
Internet Usage Dataset	General demographics of internet users.	None.	10,104	Text	Classification, clustering	1999		D. Cook
URL Dataset	120 days of URL data from a large conference.	Many features of each URL are given.	2,396,130	Text	Classification	2009		J. Ma
Phishing Websites Dataset	Dataset of phishing websites.	Many features of each site are given.	2456	Text	Classification	2015		R. Mustafa et al.
Online Retail Dataset	Online transactions for a UK online retailer.	Details of each transaction given.	541,909	Text	Classification, clustering	2015		D. Chen
Freebase Simple Topic Dump	Freebase is an online effort to structure all human knowledge.	Topics from Freebase have been extracted.	large	Text	Classification, clustering	2011		Freebase
Farm Ads Dataset	The text of farm ads from websites. Binary approval or disapproval by content owners is given.	SVMlight sparse vectors of text words in ads calculated.	4143	Text	Classification	2011		C. Masterharm et al.

Games

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Poker Hand Dataset	5 card hands from a standard 52 card deck.	Attributes of each hand are given, including the Poker hands formed by the cards it contains.	1,025,010	Text	Regression, classification	2007		R. Cattral
Connect-4 Dataset	Contains all legal 8-ply positions in the game of connect-4 in which neither player has won yet, and in which the next move is not forced.	None.	67,557	Text	Classification	1995		J. Tromp
Chess (King-Rook vs. King) Dataset	Endgame Database for White King and Rook against Black King.	None.	28,056	Text	Classification	1994		M. Bain et al.
Chess (King-Rook vs. King-Pawn) Dataset	King+Rook versus King+Pawn on a7.	None.	3196	Text	Classification	1989		R. Holte
Tic-Tac-Toe Endgame Dataset	Binary classification for win conditions in tic-tac-toe.	None.	958	Text	Classification	1991		D. Aha

Other multivariate

Dataset Name	Brief description	Preprocessing	Instances	Format	Default Task	Created (updated)	Reference	Creator
Housing Data Set	Median home values of Boston with associated home and neighborhood attributes.	None.	506	Text	Regression	1993		D. Harrison et al.
The Getty Vocabularies	structured terminology for art and other material culture, archival materials, visual surrogates, and bibliographic materials.	None.	large	Text	Classification	2015		Getty Center
Yahoo! Front Page Today Module User Click Log	User click log for news articles displayed in the Featured Tab of the Today Module on Yahoo! Front Page.	Conjoint analysis with a bilinear model.	45,811,883 user visits	Text	Regression, clustering	2009		Chu et al.
British Oceanographic Data Centre	Biological, chemical, physical and geophysical data for oceans. 22K variables tracked.	Various.	22K variables, many instances	Text	Regression, clustering	2015		British Oceanographic Data Centre
Congressional Voting Records Dataset	Voting data for all USA representatives on 16 issues.	Beyond the raw voting data, various other features are provided.	435	Text	Classification	1987		J. Schlimmer
Entree Chicago Recommendation Dataset	Record of user interactions with Entree Chicago recommendation system.	Details of each users usage of the app are recorded in detail.	50,672	Text	Regression, recommendation	2000		R. Burke
Insurance Company Benchmark (COIL 2000)	Information on customers of an insurance company.	Many features of each customer and the services they use.	9,000	Text	Regression, classification	2000		P. van der Putten
Nursery Dataset	Data from applicants to nursery schools.	Data about applicant's family and various other factors included.	12,960	Text	Classification	1997		V. Rajkovic et al.
University Dataset	Data describing attributed of a large number of universities.	None.	285	Text	Clustering, classification	1988		S. Sounders et al.
Blood Transfusion Service Center Dataset	Data from blood transfusion service center. Gives data on donors return rate, frequency, etc.	None.	748	Text	Classification	2008		I. Yeh
Record Linkage Comparison Patterns Dataset	Large dataset of records. Task is to link relevant records together.	Blocking procedure applied to select only certain record pairs.	5,749,132	Text	Classification	2011		University of Mainz
Nomao Dataset	Nomao collects data about places from many different sources. Task is to detect items that describe the same place.	Duplicates labeled.	34,465	Text	Classification	2012		Nomao Labs
Movie Dataset	Data for 10,000 movies.	Several features for each movie are given.	10,000	Text	Clustering, classification	1999		G. Wiederhold
Open University Learning Analytics Dataset	Information about students and their interactions with a virtual learning environment.	None.	~ 30,000	Text	Classification, clustering, regression	2015		J. Kuzilek et al.
Mobile phone records	Telecommunications activity and interactions	Aggregation per geographical grid cells and every 15 minutes.	large	Text	Classification, Clustering, Regression	2015		G. Barlacchi et al.

Curated repositories of datasets

As datasets come in myriad formats and can sometimes be difficult to use, there has been considerable work put into curating and standardizing the format of datasets to make them easier to use for machine learning research.

• OpenML: Web platform with Python, R, Java, and other APIs for downloading hundreds of machine learning datasets, evaluating algorithms on datasets, and benchmarking algorithm performance against dozens of other algorithms.
• PMLB: A large, curated repository of benchmark datasets for evaluating supervised machine learning algorithms. Provides classification and regression datasets in a standardized format that are accessible through a Python API.
• Metatext NLP: https://metatext.io/datasets web repository maintained by community, containing nearly 1000 benchmark datasets, and counting. Provides many tasks from classification to QA, and various languages from English, Portuguese to Arabic.
• Appen: Off The Shelf and Open Source Datasets hosted and maintained by the company. These biological, image, physical, question answering, signal, sound, text, and video resources number over 250 and can be applied to over 25 different use cases.

See also

• Comparison of deep learning software
• List of manual image annotation tools
• List of biological databases