Dataset

This dataset was created by ShubhamSawant

Released under CC0: Public Domain

Contents

This is the second version of the Google Landmarks dataset (GLDv2), which contains images annotated with labels representing human-made and natural landmarks. The dataset can be used for landmark recognition and retrieval experiments. This version of the dataset contains approximately 5 million images, split into 3 sets of images: train, index and test. The dataset was presented in our CVPR'20 paper. In this repository, we present download links for all dataset files and relevant code for metric computation. This dataset was associated to two Kaggle challenges, on landmark recognition and landmark retrieval. Results were discussed as part of a CVPR'19 workshop. In this repository, we also provide scores for the top 10 teams in the challenges, based on the latest ground-truth version. Please visit the challenge and workshop webpages for more details on the data, tasks and technical solutions from top teams.

Context

The Google Merchandise Store sells Google branded merchandise. The data is typical of what you would see for an ecommerce website.

Content

The sample dataset contains Google Analytics 360 data from the Google Merchandise Store, a real ecommerce store. The Google Merchandise Store sells Google branded merchandise. The data is typical of what you would see for an ecommerce website. It includes the following kinds of information:

Traffic source data: information about where website visitors originate. This includes data about organic traffic, paid search traffic, display traffic, etc. Content data: information about the behavior of users on the site. This includes the URLs of pages that visitors look at, how they interact with content, etc. Transactional data: information about the transactions that occur on the Google Merchandise Store website.

Fork this kernel to get started.

Acknowledgements

Data from: https://bigquery.cloud.google.com/table/bigquery-public-data:google_analytics_sample.ga_sessions_20170801

Banner Photo by Edho Pratama from Unsplash.

Inspiration

What is the total number of transactions generated per device browser in July 2017?

The real bounce rate is defined as the percentage of visits with a single pageview. What was the real bounce rate per traffic source?

What was the average number of product pageviews for users who made a purchase in July 2017?

What was the average number of product pageviews for users who did not make a purchase in July 2017?

What was the average total transactions per user that made a purchase in July 2017?

What is the average amount of money spent per session in July 2017?

What is the sequence of pages viewed?

Dataset

This dataset was created by Zacks Shen

Contents

It contains the following files:

Images dataset divided into train (10905114 images), validation (2115528 images) and test (544946 images) folders containing a balanced number of images for two classes (chemical structures and non-chemical structures).

The chemical structures were generated using RanDepict to random picked compounds from the ChEMBL30 database and the COCONUT database.

The non-chemical structures were generated using Python or they were retrieved from several public datasets:

COCO dataset, MIT Places-205 dataset, Visual Genome dataset, Google Open labeled Images, MMU-OCR-21 (kaggle), HandWritten_Character (kaggle), CoronaHack -Chest X-Ray-dataset (kaggle), PANDAS Augmented Images (kaggle), Bacterial_Colony (kaggle), Ceylon Epigraphy Periods (kaggle), Chinese Calligraphy Styles by Calligraphers (kaggle), Graphs Dataset (kaggle), Function_Graphs Polynomial (kaggle), sketches (kaggle), Person Face Sketches (kaggle), Art Pictograms (kaggle), Russian handwritten letters (kaggle), Handwritten Russian Letters (kaggle), Covid-19 Misinformation Tweets Labeled Dataset (kaggle) and grapheme-imgs-224x224 (kaggle).

This data was used to build a CNN classification model using as a base model EfficienNetB0 and fine tuning it. The model is available on Github.

Dataset Card for Danish WIT

  Dataset Summary

Google presented the Wikipedia Image Text (WIT) dataset in July 2021, a dataset which contains scraped images from Wikipedia along with their descriptions. WikiMedia released WIT-Base in September 2021, being a modified version of WIT where they have removed the images with empty "reference descriptions", as well as removing images where a person's face covers more than 10% of the image surface, along with inappropriate… See the full description on the dataset page: https://huggingface.co/datasets/huggingface/test-big-dataset.

FSDKaggle2019 is an audio dataset containing 29,266 audio files annotated with 80 labels of the AudioSet Ontology. FSDKaggle2019 has been used for the DCASE Challenge 2019 Task 2, which was run as a Kaggle competition titled Freesound Audio Tagging 2019.

Citation

If you use the FSDKaggle2019 dataset or part of it, please cite our DCASE 2019 paper:

Eduardo Fonseca, Manoj Plakal, Frederic Font, Daniel P. W. Ellis, Xavier Serra. "Audio tagging with noisy labels and minimal supervision". Proceedings of the DCASE 2019 Workshop, NYC, US (2019)

You can also consider citing our ISMIR 2017 paper, which describes how we gathered the manual annotations included in FSDKaggle2019.

Eduardo Fonseca, Jordi Pons, Xavier Favory, Frederic Font, Dmitry Bogdanov, Andres Ferraro, Sergio Oramas, Alastair Porter, and Xavier Serra, "Freesound Datasets: A Platform for the Creation of Open Audio Datasets", In Proceedings of the 18th International Society for Music Information Retrieval Conference, Suzhou, China, 2017

Data curators

Eduardo Fonseca, Manoj Plakal, Xavier Favory, Jordi Pons

Contact

You are welcome to contact Eduardo Fonseca should you have any questions at eduardo.fonseca@upf.edu.

ABOUT FSDKaggle2019

Freesound Dataset Kaggle 2019 (or FSDKaggle2019 for short) is an audio dataset containing 29,266 audio files annotated with 80 labels of the AudioSet Ontology [1]. FSDKaggle2019 has been used for the Task 2 of the Detection and Classification of Acoustic Scenes and Events (DCASE) Challenge 2019. Please visit the DCASE2019 Challenge Task 2 website for more information. This Task was hosted on the Kaggle platform as a competition titled Freesound Audio Tagging 2019. It was organized by researchers from the Music Technology Group (MTG) of Universitat Pompeu Fabra (UPF), and from Sound Understanding team at Google AI Perception. The competition intended to provide insight towards the development of broadly-applicable sound event classifiers able to cope with label noise and minimal supervision conditions.

FSDKaggle2019 employs audio clips from the following sources:

Freesound Dataset (FSD): a dataset being collected at the MTG-UPF based on Freesound content organized with the AudioSet Ontology

The soundtracks of a pool of Flickr videos taken from the Yahoo Flickr Creative Commons 100M dataset (YFCC)

The audio data is labeled using a vocabulary of 80 labels from Google’s AudioSet Ontology [1], covering diverse topics: Guitar and other Musical Instruments, Percussion, Water, Digestive, Respiratory sounds, Human voice, Human locomotion, Hands, Human group actions, Insect, Domestic animals, Glass, Liquid, Motor vehicle (road), Mechanisms, Doors, and a variety of Domestic sounds. The full list of categories can be inspected in vocabulary.csv (see Files & Download below). The goal of the task was to build a multi-label audio tagging system that can predict appropriate label(s) for each audio clip in a test set.

What follows is a summary of some of the most relevant characteristics of FSDKaggle2019. Nevertheless, it is highly recommended to read our DCASE 2019 paper for a more in-depth description of the dataset and how it was built.

Ground Truth Labels

The ground truth labels are provided at the clip-level, and express the presence of a sound category in the audio clip, hence can be considered weak labels or tags. Audio clips have variable lengths (roughly from 0.3 to 30s).

The audio content from FSD has been manually labeled by humans following a data labeling process using the Freesound Annotator platform. Most labels have inter-annotator agreement but not all of them. More details about the data labeling process and the Freesound Annotator can be found in [2].

The YFCC soundtracks were labeled using automated heuristics applied to the audio content and metadata of the original Flickr clips. Hence, a substantial amount of label noise can be expected. The label noise can vary widely in amount and type depending on the category, including in- and out-of-vocabulary noises. More information about some of the types of label noise that can be encountered is available in [3].

Specifically, FSDKaggle2019 features three types of label quality, one for each set in the dataset:

curated train set: correct (but potentially incomplete) labels

noisy train set: noisy labels

test set: correct and complete labels

Further details can be found below in the sections for each set.

Format

All audio clips are provided as uncompressed PCM 16 bit, 44.1 kHz, mono audio files.

DATA SPLIT

FSDKaggle2019 consists of two train sets and one test set. The idea is to limit the supervision provided for training (i.e., the manually-labeled, hence reliable, data), thus promoting approaches to deal with label noise.

Curated train set

The curated train set consists of manually-labeled data from FSD.

Number of clips/class: 75 except in a few cases (where there are less)

Total number of clips: 4970

Avg number of labels/clip: 1.2

Total duration: 10.5 hours

The duration of the audio clips ranges from 0.3 to 30s due to the diversity of the sound categories and the preferences of Freesound users when recording/uploading sounds. Labels are correct but potentially incomplete. It can happen that a few of these audio clips present additional acoustic material beyond the provided ground truth label(s).

Noisy train set

The noisy train set is a larger set of noisy web audio data from Flickr videos taken from the YFCC dataset [5].

Number of clips/class: 300

Total number of clips: 19,815

Avg number of labels/clip: 1.2

Total duration: ~80 hours

The duration of the audio clips ranges from 1s to 15s, with the vast majority lasting 15s. Labels are automatically generated and purposefully noisy. No human validation is involved. The label noise can vary widely in amount and type depending on the category, including in- and out-of-vocabulary noises.

Considering the numbers above, the per-class data distribution available for training is, for most of the classes, 300 clips from the noisy train set and 75 clips from the curated train set. This means 80% noisy / 20% curated at the clip level, while at the duration level the proportion is more extreme considering the variable-length clips.

Test set

The test set is used for system evaluation and consists of manually-labeled data from FSD.

Number of clips/class: between 50 and 150

Total number of clips: 4481

Avg number of labels/clip: 1.4

Total duration: 12.9 hours

The acoustic material present in the test set clips is labeled exhaustively using the aforementioned vocabulary of 80 classes. Most labels have inter-annotator agreement but not all of them. Except human error, the label(s) are correct and complete considering the target vocabulary; nonetheless, a few clips could still present additional (unlabeled) acoustic content out of the vocabulary.

During the DCASE2019 Challenge Task 2, the test set was split into two subsets, for the public and private leaderboards, and only the data corresponding to the public leaderboard was provided. In this current package you will find the full test set with all the test labels. To allow comparison with previous work, the file test_post_competition.csv includes a flag to determine the corresponding leaderboard (public or private) for each test clip (see more info in Files & Download below).

Acoustic mismatch

As mentioned before, FSDKaggle2019 uses audio clips from two sources:

FSD: curated train set and test set, and

YFCC: noisy train set.

While the sources of audio (Freesound and Flickr) are collaboratively contributed and pretty diverse themselves, a certain acoustic mismatch can be expected between FSD and YFCC. We conjecture this mismatch comes from a variety of reasons. For example, through acoustic inspection of a small sample of both data sources, we find a higher percentage of high quality recordings in FSD. In addition, audio clips in Freesound are typically recorded with the purpose of capturing audio, which is not necessarily the case in YFCC.

This mismatch can have an impact in the evaluation, considering that most of the train data come from YFCC, while all test data are drawn from FSD. This constraint (i.e., noisy training data coming from a different web audio source than the test set) is sometimes a real-world condition.

LICENSE

All clips in FSDKaggle2019 are released under Creative Commons (CC) licenses. For attribution purposes and to facilitate attribution of these files to third parties, we include a mapping from the audio clips to their corresponding licenses.

Curated train set and test set. All clips in Freesound are released under different modalities of Creative Commons (CC) licenses, and each audio clip has its own license as defined by the audio clip uploader in Freesound, some of them requiring attribution to their original authors and some forbidding further commercial reuse. The licenses are specified in the files train_curated_post_competition.csv and test_post_competition.csv. These licenses can be CC0, CC-BY, CC-BY-NC and CC Sampling+.

Noisy train set. Similarly, the licenses of the soundtracks from Flickr used in FSDKaggle2019 are specified in the file train_noisy_post_competition.csv. These licenses can be CC-BY and CC BY-SA.

In addition, FSDKaggle2019 as a whole is the result of a curation process and it has an additional license. FSDKaggle2019 is released under CC-BY. This license is specified in the LICENSE-DATASET file downloaded with the FSDKaggle2019.doc zip file.

FILES & DOWNLOAD

FSDKaggle2019 can be downloaded as a series of zip files with the following directory structure:

root │
└───FSDKaggle2019.audio_train_curated/ Audio clips in the curated train set │ └───FSDKaggle2019.audio_train_noisy/ Audio clips in the noisy

SAT History Questions and Answers 🏛️ - Text Classification Dataset

This dataset contains a collection of questions and answers for the SAT Subject Test in World History and US History. Each question is accompanied by a corresponding answers and the correct response.

The dataset includes questions from various topics, time periods, and regions on both World History and US History.

💴 For Commercial Usage: To discuss your requirements, learn about the price and buy the dataset, leave a request on TrainingData to buy the dataset

OTHER DATASETS FOR THE TEXT ANALYSIS:

• Google Play Messengers - 6,000 Reviews ⭐️
• 20,000 Customers Reviews on Banks ⭐️
• Amazon Reviews Dataset

Content

For each question, we extracted: - id: number of the question, - subject: SAT subject (World History or US History), - prompt: text of the question, - A: answer A, - B: answer B, - C: answer C, - D: answer D, - E: answer E, - answer: letter of the correct answer to the question

💴 Buy the Dataset: This is just an example of the data. Leave a request on https://trainingdata.pro/datasets to discuss your requirements, learn about the price and buy the dataset

TrainingData provides high-quality data annotation tailored to your needs

keywords: answer questions, sat, gpa, university, school, exam, college, web scraping, parsing, online database, text dataset, sentiment analysis, llm dataset, language modeling, large language models, text classification, text mining dataset, natural language texts, nlp, nlp open-source dataset, text data, machine learning

FSDKaggle2018 is an audio dataset containing 11,073 audio files annotated with 41 labels of the AudioSet Ontology. FSDKaggle2018 has been used for the DCASE Challenge 2018 Task 2, which was run as a Kaggle competition titled Freesound General-Purpose Audio Tagging Challenge.

Citation

If you use the FSDKaggle2018 dataset or part of it, please cite our DCASE 2018 paper:

Eduardo Fonseca, Manoj Plakal, Frederic Font, Daniel P. W. Ellis, Xavier Favory, Jordi Pons, Xavier Serra. "General-purpose Tagging of Freesound Audio with AudioSet Labels: Task Description, Dataset, and Baseline". Proceedings of the DCASE 2018 Workshop (2018)

You can also consider citing our ISMIR 2017 paper, which describes how we gathered the manual annotations included in FSDKaggle2018.

Eduardo Fonseca, Jordi Pons, Xavier Favory, Frederic Font, Dmitry Bogdanov, Andres Ferraro, Sergio Oramas, Alastair Porter, and Xavier Serra, "Freesound Datasets: A Platform for the Creation of Open Audio Datasets", In Proceedings of the 18th International Society for Music Information Retrieval Conference, Suzhou, China, 2017

Contact

You are welcome to contact Eduardo Fonseca should you have any questions at eduardo.fonseca@upf.edu.

About this dataset

Freesound Dataset Kaggle 2018 (or FSDKaggle2018 for short) is an audio dataset containing 11,073 audio files annotated with 41 labels of the AudioSet Ontology [1]. FSDKaggle2018 has been used for the Task 2 of the Detection and Classification of Acoustic Scenes and Events (DCASE) Challenge 2018. Please visit the DCASE2018 Challenge Task 2 website for more information. This Task was hosted on the Kaggle platform as a competition titled Freesound General-Purpose Audio Tagging Challenge. It was organized by researchers from the Music Technology Group of Universitat Pompeu Fabra, and from Google Research’s Machine Perception Team.

The goal of this competition was to build an audio tagging system that can categorize an audio clip as belonging to one of a set of 41 diverse categories drawn from the AudioSet Ontology.

All audio samples in this dataset are gathered from Freesound [2] and are provided here as uncompressed PCM 16 bit, 44.1 kHz, mono audio files. Note that because Freesound content is collaboratively contributed, recording quality and techniques can vary widely.

The ground truth data provided in this dataset has been obtained after a data labeling process which is described below in the Data labeling process section. FSDKaggle2018 clips are unequally distributed in the following 41 categories of the AudioSet Ontology:

"Acoustic_guitar", "Applause", "Bark", "Bass_drum", "Burping_or_eructation", "Bus", "Cello", "Chime", "Clarinet", "Computer_keyboard", "Cough", "Cowbell", "Double_bass", "Drawer_open_or_close", "Electric_piano", "Fart", "Finger_snapping", "Fireworks", "Flute", "Glockenspiel", "Gong", "Gunshot_or_gunfire", "Harmonica", "Hi-hat", "Keys_jangling", "Knock", "Laughter", "Meow", "Microwave_oven", "Oboe", "Saxophone", "Scissors", "Shatter", "Snare_drum", "Squeak", "Tambourine", "Tearing", "Telephone", "Trumpet", "Violin_or_fiddle", "Writing".

Some other relevant characteristics of FSDKaggle2018:

• The dataset is split into a train set and a test set.

• The train set is meant to be for system development and includes ~9.5k samples unequally distributed among 41 categories. The minimum number of audio samples per category in the train set is 94, and the maximum 300. The duration of the audio samples ranges from 300ms to 30s due to the diversity of the sound categories and the preferences of Freesound users when recording sounds. The total duration of the train set is roughly 18h.

• Out of the ~9.5k samples from the train set, ~3.7k have manually-verified ground truth annotations and ~5.8k have non-verified annotations. The non-verified annotations of the train set have a quality estimate of at least 65-70% in each category. Checkout the Data labeling process section below for more information about this aspect.

• Non-verified annotations in the train set are properly flagged in train.csv so that participants can opt to use this information during the development of their systems.

• The test set is composed of 1.6k samples with manually-verified annotations and with a similar category distribution than that of the train set. The total duration of the test set is roughly 2h.

• All audio samples in this dataset have a single label (i.e. are only annotated with one label). Checkout the Data labeling process section below for more information about this aspect. A single label should be predicted for each file in the test set.

Data labeling process

The data labeling process started from a manual mapping between Freesound tags and AudioSet Ontology categories (or labels), which was carried out by researchers at the Music Technology Group, Universitat Pompeu Fabra, Barcelona. Using this mapping, a number of Freesound audio samples were automatically annotated with labels from the AudioSet Ontology. These annotations can be understood as weak labels since they express the presence of a sound category in an audio sample.

Then, a data validation process was carried out in which a number of participants did listen to the annotated sounds and manually assessed the presence/absence of an automatically assigned sound category, according to the AudioSet category description.

Audio samples in FSDKaggle2018 are only annotated with a single ground truth label (see train.csv). A total of 3,710 annotations included in the train set of FSDKaggle2018 are annotations that have been manually validated as present and predominant (some with inter-annotator agreement but not all of them). This means that in most cases there is no additional acoustic material other than the labeled category. In few cases there may be some additional sound events, but these additional events won't belong to any of the 41 categories of FSDKaggle2018.

The rest of the annotations have not been manually validated and therefore some of them could be inaccurate. Nonetheless, we have estimated that at least 65-70% of the non-verified annotations per category in the train set are indeed correct. It can happen that some of these non-verified audio samples present several sound sources even though only one label is provided as ground truth. These additional sources are typically out of the set of the 41 categories, but in a few cases they could be within.

More details about the data labeling process can be found in [3].

License

FSDKaggle2018 has licenses at two different levels, as explained next.

All sounds in Freesound are released under Creative Commons (CC) licenses, and each audio clip has its own license as defined by the audio clip uploader in Freesound. For attribution purposes and to facilitate attribution of these files to third parties, we include a relation of the audio clips included in FSDKaggle2018 and their corresponding license. The licenses are specified in the files train_post_competition.csv and test_post_competition_scoring_clips.csv.

In addition, FSDKaggle2018 as a whole is the result of a curation process and it has an additional license. FSDKaggle2018 is released under CC-BY. This license is specified in the LICENSE-DATASET file downloaded with the FSDKaggle2018.doc zip file.

Files

FSDKaggle2018 can be downloaded as a series of zip files with the following directory structure:

root
│

└───FSDKaggle2018.audio_train/            Audio clips in the train set
│

└───FSDKaggle2018.audio_test/            Audio clips in the test set
│

└───FSDKaggle2018.meta/               Files for evaluation setup
│  │

│  └───train_post_competition.csv          Data split and ground truth for the train set
│  │

│  └───test_post_competition_scoring_clips.csv   Ground truth for the test set

│

└───FSDKaggle2018.doc/
  │

  └───README.md                  The dataset description file you are reading
  │

  └───LICENSE-DATASET             

This noisy speech test set is created from the Google Speech Commands v2 [1] and the Musan dataset[2].

It could be downloaded here: https://zenodo.org/record/6066174#.Yn7NPJPMLyU

Specifically, we created this test set by mixing the speech in the Google Speech Commands v2 test set with random noise in the Musan dataset at different signal to noise ratio -12.5,-10,0,10,20,30 and 40 decibel (dB).

The Google Speech Commands v2 dataset is under the Creative Commons BY 4.0 license. It could be downloaded at: http://download.tensorflow.org/data/speech_commands_v0.02.tar.gz

The Musan dataset is under Attribution 4.0 International (CC BY 4.0). It could be downlowned at https://www.openslr.org/17/

Citations:

[1] Pete Warden, “Speech commands: A dataset for limited-vocabulary speech recognition,” arXiv preprint arXiv:1804.03209, 2018.

[2] David Snyder, Guoguo Chen, and Daniel Povey, “Musan: A music, speech, and noise corpus,” arXiv preprint arXiv:1510.08484, 2015.

Dataset Card for Boolq

  Dataset Summary

BoolQ is a question answering dataset for yes/no questions containing 15942 examples. These questions are naturally occurring ---they are generated in unprompted and unconstrained settings. Each example is a triplet of (question, passage, answer), with the title of the page as optional additional context. The text-pair classification setup is similar to existing natural language inference tasks.

  Supported Tasks… See the full description on the dataset page: https://huggingface.co/datasets/google/boolq.

Overview

This dataset of medical misinformation was collected and is published by Kempelen Institute of Intelligent Technologies (KInIT). It consists of approx. 317k news articles and blog posts on medical topics published between January 1, 1998 and February 1, 2022 from a total of 207 reliable and unreliable sources. The dataset contains full-texts of the articles, their original source URL and other extracted metadata. If a source has a credibility score available (e.g., from Media Bias/Fact Check), it is also included in the form of annotation. Besides the articles, the dataset contains around 3.5k fact-checks and extracted verified medical claims with their unified veracity ratings published by fact-checking organisations such as Snopes or FullFact. Lastly and most importantly, the dataset contains 573 manually and more than 51k automatically labelled mappings between previously verified claims and the articles; mappings consist of two values: claim presence (i.e., whether a claim is contained in the given article) and article stance (i.e., whether the given article supports or rejects the claim or provides both sides of the argument).

The dataset is primarily intended to be used as a training and evaluation set for machine learning methods for claim presence detection and article stance classification, but it enables a range of other misinformation related tasks, such as misinformation characterisation or analyses of misinformation spreading.

Its novelty and our main contributions lie in (1) focus on medical news article and blog posts as opposed to social media posts or political discussions; (2) providing multiple modalities (beside full-texts of the articles, there are also images and videos), thus enabling research of multimodal approaches; (3) mapping of the articles to the fact-checked claims (with manual as well as predicted labels); (4) providing source credibility labels for 95% of all articles and other potential sources of weak labels that can be mined from the articles' content and metadata.

The dataset is associated with the research paper "Monant Medical Misinformation Dataset: Mapping Articles to Fact-Checked Claims" accepted and presented at ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR '22).

The accompanying Github repository provides a small static sample of the dataset and the dataset's descriptive analysis in a form of Jupyter notebooks.

Options to access the dataset

There are two ways how to get access to the dataset:

1. Static dump of the dataset available in the CSV format
2. Continuously updated dataset available via REST API

In order to obtain an access to the dataset (either to full static dump or REST API), please, request the access by following instructions provided below.

References

If you use this dataset in any publication, project, tool or in any other form, please, cite the following papers:

@inproceedings{SrbaMonantPlatform,
  author = {Srba, Ivan and Moro, Robert and Simko, Jakub and Sevcech, Jakub and Chuda, Daniela and Navrat, Pavol and Bielikova, Maria},
  booktitle = {Proceedings of Workshop on Reducing Online Misinformation Exposure (ROME 2019)},
  pages = {1--7},
  title = {Monant: Universal and Extensible Platform for Monitoring, Detection and Mitigation of Antisocial Behavior},
  year = {2019}
}

@inproceedings{SrbaMonantMedicalDataset,
  author = {Srba, Ivan and Pecher, Branislav and Tomlein Matus and Moro, Robert and Stefancova, Elena and Simko, Jakub and Bielikova, Maria},
  booktitle = {Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR '22)},
  numpages = {11},
  title = {Monant Medical Misinformation Dataset: Mapping Articles to Fact-Checked Claims},
  year = {2022},
  doi = {10.1145/3477495.3531726},
  publisher = {Association for Computing Machinery},
  address = {New York, NY, USA},
  url = {https://doi.org/10.1145/3477495.3531726},
}

Dataset creation process

In order to create this dataset (and to continuously obtain new data), we used our research platform Monant. The Monant platform provides so called data providers to extract news articles/blogs from news/blog sites as well as fact-checking articles from fact-checking sites. General parsers (from RSS feeds, Wordpress sites, Google Fact Check Tool, etc.) as well as custom crawler and parsers were implemented (e.g., for fact checking site Snopes.com). All data is stored in the unified format in a central data storage.


Ethical considerations

The dataset was collected and is published for research purposes only. We collected only publicly available content of news/blog articles. The dataset contains identities of authors of the articles if they were stated in the original source; we left this information, since the presence of an author's name can be a strong credibility indicator. However, we anonymised the identities of the authors of discussion posts included in the dataset.

The main identified ethical issue related to the presented dataset lies in the risk of mislabelling of an article as supporting a false fact-checked claim and, to a lesser extent, in mislabelling an article as not containing a false claim or not supporting it when it actually does. To minimise these risks, we developed a labelling methodology and require an agreement of at least two independent annotators to assign a claim presence or article stance label to an article. It is also worth noting that we do not label an article as a whole as false or true. Nevertheless, we provide partial article-claim pair veracities based on the combination of claim presence and article stance labels.

As to the veracity labels of the fact-checked claims and the credibility (reliability) labels of the articles' sources, we take these from the fact-checking sites and external listings such as Media Bias/Fact Check as they are and refer to their methodologies for more details on how they were established.

Lastly, the dataset also contains automatically predicted labels of claim presence and article stance using our baselines described in the next section. These methods have their limitations and work with certain accuracy as reported in this paper. This should be taken into account when interpreting them.


Reporting mistakes in the dataset
The mean to report considerable mistakes in raw collected data or in manual annotations is by creating a new issue in the accompanying Github repository. Alternately, general enquiries or requests can be sent at info [at] kinit.sk.

Dataset structure

Raw data

At first, the dataset contains so called raw data (i.e., data extracted by the Web monitoring module of Monant platform and stored in exactly the same form as they appear at the original websites). Raw data consist of articles from news sites and blogs (e.g. naturalnews.com), discussions attached to such articles, fact-checking articles from fact-checking portals (e.g. snopes.com). In addition, the dataset contains feedback (number of likes, shares, comments) provided by user on social network Facebook which is regularly extracted for all news/blogs articles.

Raw data are contained in these CSV files (and corresponding REST API endpoints):

• sources.csv
• articles.csv
• article_media.csv
• article_authors.csv
• discussion_posts.csv
• discussion_post_authors.csv
• fact_checking_articles.csv
• fact_checking_article_media.csv
• claims.csv
• feedback_facebook.csv

Note: Personal information about discussion posts' authors (name, website, gravatar) are anonymised.

Annotations

Secondly, the dataset contains so called annotations. Entity annotations describe the individual raw data entities (e.g., article, source). Relation annotations describe relation between two of such entities.

Each annotation is described by the following attributes:

1. category of annotation (`annotation_category`). Possible values: label (annotation corresponds to ground truth, determined by human experts) and prediction (annotation was created by means of AI method).
2. type of annotation (`annotation_type_id`). Example values: Source reliability (binary), Claim presence. The list of possible values can be obtained from enumeration in annotation_types.csv.
3. method which created annotation (`method_id`). Example values: Expert-based source reliability evaluation, Fact-checking article to claim transformation method. The list of possible values can be obtained from enumeration methods.csv.
4. its value (`value`). The value is stored in JSON format and its structure differs according to particular annotation type.

At the same time, annotations are associated with a particular object identified by:

1. entity type (parameter entity_type in case of entity annotations, or source_entity_type and target_entity_type in case of relation annotations). Possible values: sources, articles, fact-checking-articles.
2. entity id (parameter entity_id in case of entity annotations, or source_entity_id and target_entity_id in case of relation

We developed a pix2pix deep learning model for segmentation of subretinal fluid area in fundus photographs to detect central serous chorioretinopathy (CSC).

The dataset include fundus photographs and segmentation images from 105 eyes with CSC and 40 healthy eyes. We retrospectively reviewed the medical records and multimodal images of a total of 115 images of patients with had CSC at Aerospace Medical Center and from publicly accessible databases. Finally, the total dataset includes fundus photographs and segmentation images from 115 eyes with CSC and 40 healthy eyes from the medical center and publicly accessible datasets. The reference segmentation for subretinal fluid area was performed manually by an expert ophthalmologist.

First, the user should upload "pix2pix_csc_segmentation.ipynb" file in the Google drive. And open the file in the Google drive page. Second, please link the datasets to this colab notebook using Google drive. For example, we save the training dataset at "csc/segmentation/seg_pix/" (in the example.zip file) and the test dataset at "csc/segmentation/seg_pix_test/" (in the example.zip file, too). Third, run the codes in Google Colab by clicking buttons.

Corners Test

## Overview

Corners Test is a dataset for object detection tasks - it contains Corners Test annotations for 207 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

Consists of 37,723/14,360 person/clothes images, with the resolution of 256x192. Each person has different poses. We split them into the train/test set 52,236/10,544 three-tuples, respectively. You can download the dataset at MPV(Google Drive)

Database description:

The written and spoken digits database is not a new database but a constructed database from existing ones, in order to provide a ready-to-use database for multimodal fusion [1].

The written digits database is the original MNIST handwritten digits database [2] with no additional processing. It consists of 70000 images (60000 for training and 10000 for test) of 28 x 28 = 784 dimensions.

The spoken digits database was extracted from Google Speech Commands [3], an audio dataset of spoken words that was proposed to train and evaluate keyword spotting systems. It consists of 105829 utterances of 35 words, amongst which 38908 utterances of the ten digits (34801 for training and 4107 for test). A pre-processing was done via the extraction of the Mel Frequency Cepstral Coefficients (MFCC) with a framing window size of 50 ms and frame shift size of 25 ms. Since the speech samples are approximately 1 s long, we end up with 39 time slots. For each one, we extract 12 MFCC coefficients with an additional energy coefficient. Thus, we have a final vector of 39 x 13 = 507 dimensions. Standardization and normalization were applied on the MFCC features.

To construct the multimodal digits dataset, we associated written and spoken digits of the same class respecting the initial partitioning in [2] and [3] for the training and test subsets. Since we have less samples for the spoken digits, we duplicated some random samples to match the number of written digits and have a multimodal digits database of 70000 samples (60000 for training and 10000 for test).

The dataset is provided in six files as described below. Therefore, if a shuffle is performed on the training or test subsets, it must be performed in unison with the same order for the written digits, spoken digits and labels.

Files:

• data_wr_train.npy: 60000 samples of 784-dimentional written digits for training;
• data_sp_train.npy: 60000 samples of 507-dimentional spoken digits for training;
• labels_train.npy: 60000 labels for the training subset;
• data_wr_test.npy: 10000 samples of 784-dimentional written digits for test;
• data_sp_test.npy: 10000 samples of 507-dimentional spoken digits for test;
• labels_test.npy: 10000 labels for the test subset.

References:

1. Khacef, L. et al. (2020), "Brain-Inspired Self-Organization with Cellular Neuromorphic Computing for Multimodal Unsupervised Learning".
2. LeCun, Y. & Cortes, C. (1998), “MNIST handwritten digit database”.
3. Warden, P. (2018), “Speech Commands: A Dataset for Limited-Vocabulary Speech Recognition”.

Context

Images of small objects for small instance detections. Currently, four object types are available.

Content

We collect four datasets of small objects from images/videos on the Internet (e.g.YouTube or Google).

Fly Dataset: contains 600 video frames with an average of 86 ± 39 flies per frame (648×72 @ 30 fps). 32 images are used for training (1:6:187) and 50 images for testing (301:6:600).

Honeybee Dataset: contains 118 images with an average of 28 ± 6 honeybees per image (640×480). The dataset is divided evenly for training and test sets. Only the first 32 images are used for training.

Fish Dataset: contains 387 frames of video with an average of 56±9 fish per frame (300×410 @ 30 fps). 32 images are used for training (1:3:94) and 65 for testing (193:3:387).

Seagull Dataset: contains three high-resolution images (624×964) with an average of 866±107 seagulls per image. The first image is used for training, and the rest for testing.

Acknowledgements

Citation: Small Instance Detection by Integer Programming on Object Density Maps. Zheng Ma, Lei Yu, and Antoni B. Chan, In: IEEE Conf. Computer Vision and Pattern Recognition (CVPR), Boston, Jun 2015

original form of dataset available here

Inspiration

Developing object detection algorithms that are more accurate at detecting small objects or small instances of objects.

US state level forecasts.

Dataset Card for CodeContests

  Dataset Summary

CodeContests is a competitive programming dataset for machine-learning. This dataset was used when training AlphaCode. It consists of programming problems, from a variety of sources:

Site URL Source

Aizu https://judge.u-aizu.ac.jp CodeNet

AtCoder https://atcoder.jp CodeNet

CodeChef https://www.codechef.com description2code

Codeforces https://codeforces.com description2code and Codeforces

HackerEarth… See the full description on the dataset page: https://huggingface.co/datasets/deepmind/code_contests.

The SCUT-CTW1500 dataset contains 1,500 images: 1,000 for training and 500 for testing. In particular, it provides 10,751 cropped text instance images, including 3,530 with curved text. The images are manually harvested from the Internet, image libraries such as Google Open-Image, or phone cameras. The dataset contains a lot of horizontal and multi-oriented text.