Skip to content

Latest commit

 

History

History
307 lines (151 loc) · 20.5 KB

DATASHEET.md

File metadata and controls

307 lines (151 loc) · 20.5 KB

Datasheet: VTC

Authors: Laura Hanu, James Thewlis, Yuki M. Asano, Christian Rupprecht

Organizations: Unitary, University of Amsterdam, University of Oxford

Motivation

The questions in this section are primarily intended to encourage dataset creators to clearly articulate their reasons for creating the dataset and to promote transparency about funding interests.

  1. For what purpose was the dataset created? Was there a specific task in mind? Was there a specific gap that needed to be filled? Please provide a description.

    The dataset was created strictly for research purposes. More specifically, this dataset addresses the research problem of using a weakly informative modality (user comments) in conjunction with other learning signals such as titles and videos for learning multi-modal representations.

  2. Who created this dataset (e.g. which team, research group) and on behalf of which entity (e.g. company, institution, organization)?

    This dataset is created by VGG, a research group at the University of Oxford and Unitary AI, a company that's developing AI to automate content moderation.

  3. What support was needed to make this dataset? (e.g. who funded the creation of the dataset? If there is an associated grant, provide the name of the grantor and the grant name and number, or if it was supported by a company or government agency, give those details.)

    The creation of dataset has not been funded directly. The individual researchers are funded by Amazon Machine Learning Awards (MLRA) and Innovate UK (project 71653) on behalf of UK Research and Innovation (UKRI).

  4. Any other comments?

    No.

Composition

Dataset creators should read through the questions in this section prior to any data collection and then provide answers once collection is complete. Most of these questions are intended to provide dataset consumers with the information they need to make informed decisions about using the dataset for specific tasks. The answers to some of these questions reveal information about compliance with the EU’s General Data Protection Regulation (GDPR) or comparable regulations in other jurisdictions.

  1. What do the instances that comprise the dataset represent (e.g. documents, photos, people, countries)? Are there multiple types of instances (e.g. movies, users, and ratings; people and interactions between them; nodes and edges)? Please provide a description.

    The dataset is comprised of links to videos, titles, and comments. Each video-title pair corresponds to a post on reddit.com. The dataset we share does not contain the data itself but hyperlinks to the data.

  2. How many instances are there in total (of each type, if appropriate)?

    There are 339k video-title pairs with an average of 14 comments per video.

  3. Does the dataset contain all possible instances or is it a sample (not necessarily random) of instances from a larger set? If the dataset is a sample, then what is the larger set? Is the sample representative of the larger set (e.g. geographic coverage)? If so, please describe how this representativeness was validated/verified. If it is not representative of the larger set, please describe why not (e.g. to cover a more diverse range of instances, because instances were withheld or unavailable).

    This dataset is a sample of a larger, unfiltered version of the original dataset that we have collected. From the initial version, we handpicked a list of "safe" subreddits and removed posts if: 1) they had the "NSFW" or "over_18" tags; 2) the videos contained faces or the captions contained toxic or offensive text.

  4. What data does each instance consist of? "Raw" data (e.g. unprocessed text or images) or features? In either case, please provide a description.

    Each instance consists of:

    • "reddit_id"
    • "post_url"
    • "comment_ids"
    • "subreddit"
    • "video_length"
  5. Is there a label or target associated with each instance? If so, please provide a description.

    No, there are no labels provided.

  6. Is any information missing from individual instances? If so, please provide a description, explaining why this information is missing (e.g. because it was unavailable). This does not include intentionally removed information, but might include, e.g. redacted text.

    If a user decides to remove a post, the link to the post will become invalid and thus not accessible anymore.

  7. Are relationships between individual instances made explicit (e.g. users' movie ratings, social network links)? If so, please describe how these relationships are made explicit.

    Instances that have the same subreddit are likely to share semantic meaning.

  8. Are there recommended data splits (e.g. training, development/validation, testing)? If so, please provide a description of these splits, explaining the rationale behind them.

    We will release the data splits we have used in our experiments with our code.

  9. Are there any errors, sources of noise, or redundancies in the dataset? If so, please provide a description.

    Although we have tried to remove most bot-generated text, it is likely that some noise will still exist due to the nature of this data. Similarly, a small proportion of posts might still contain identical or highly similar videos post-deduplication.

  10. Is the dataset self-contained, or does it link to or otherwise rely on external resources (e.g. websites, tweets, other datasets)? If it links to or relies on external resources, a) are there guarantees that they will exist, and remain constant, over time; b) are there official archival versions of the complete dataset (i.e., including the external resources as they existed at the time the dataset was created); c) are there any restrictions (e.g. licenses, fees) associated with any of the external resources that might apply to a future user? Please provide descriptions of all external resources and any restrictions associated with them, as well as links or other access points, as appropriate.

    If it links to or relies on external resources, a) are there guarantees that they will exist, and remain constant, over time; b) are there official archival versions of the complete dataset (i.e., including the external resources as they existed at the time the dataset was created); c) are there any restrictions (e.g., licenses, fees) associated with any of the external resources that might apply to a future user? Please provide descriptions of all external resources and any restrictions associated with them, as well as links or other access points, as appropriate.

    In order to preserve user privacy, this dataset relies on links to reddit posts and comment ids.

    a) The links will no longer be valid if a user decides to delete their post.

    b) It would be possible to find the metadata of each post, as well as the link to the media file, on the Reddit archive.

    c) All links are accessible to everyone and are likely to remain so in the future.

  11. Does the dataset contain data that might be considered confidential (e.g. data that is protected by legal privilege or by doctor-patient confidentiality, data that includes the content of individuals' non-public communications)? If so, please provide a description.

    No, all data shared links to public posts.

  12. Does the dataset contain data that, if viewed directly, might be offensive, insulting, threatening, or might otherwise cause anxiety? If so, please describe why.

    The dataset is still likely to contain a small proportion of offensive data. Due to the size of the dataset, we were not able to verify each video and each comment manually. However, we have tried to minimize the number of unsafe posts by:

    • manually filtering the subreddits included;
    • using Reddit metadata such as the "NSFW" and "over_18" tags to remove unsafe posts;
    • using automatic machine learning models to remove posts containing faces and toxic text.
  13. Does the dataset relate to people? If not, you may skip the remaining questions in this section.

    The dataset relates to people in the sense that each post is created by a person. In order to minimise the content related to people, we used a public face detector model to remove most instances of videos containing faces.

  14. Does the dataset identify any subpopulations (e.g. by age, gender)? If so, please describe how these subpopulations are identified and provide a description of their respective distributions within the dataset.

    The dataset does not explicitly identify any subpopulations. However, some titles, user comments or image contents may identify individuals as part of a subpopulation.

  15. Is it possible to identify individuals (i.e., one or more natural persons), either directly or indirectly (i.e., in combination with other data) from the dataset? If so, please describe how.

    Yes. Our dataset contains links to posts where the Reddit username will be visible and some of them might have identifying information contained in their profile such as personal images or information. This information is, however, already publicly available on Reddit.

  16. Does the dataset contain data that might be considered sensitive in any way (e.g. data that reveals racial or ethnic origins, sexual orientations, religious beliefs, political opinions or union memberships, or locations; financial or health data; biometric or genetic data; forms of government identification, such as social security numbers; criminal history)? If so, please provide a description.

    While we believe this is highly unlikely (as we only use already public posts and comments) -- we cannot rule this out with absolute certainty. We will actively maintain this dataset after its release and ensure that if such information is included, that it is removed swiftly.

  17. Any other comments?

    No.

Collection

As with the previous section, dataset creators should read through these questions prior to any data collection to flag potential issues and then provide answers once collection is complete. In addition to the goals of the prior section, the answers to questions here may provide information that allow others to reconstruct the dataset without access to it.

  1. How was the data associated with each instance acquired? Was the data directly observable (e.g. raw text, movie ratings), reported by subjects (e.g. survey responses), or indirectly inferred/derived from other data (e.g. part-of-speech tags, model-based guesses for age or language)? If data was reported by subjects or indirectly inferred/derived from other data, was the data validated/verified? If so, please describe how.

    The data was already available on Reddit.

  2. What mechanisms or procedures were used to collect the data (e.g. hardware apparatus or sensor, manual human curation, software program, software API)? How were these mechanisms or procedures validated?

    The dataset was collected via Reddit's own API.

  3. If the dataset is a sample from a larger set, what was the sampling strategy (e.g. deterministic, probabilistic with specific sampling probabilities)?

    NA.

  4. Who was involved in the data collection process (e.g. students, crowdworkers, contractors) and how were they compensated (e.g. how much were crowdworkers paid)?

    NA.

  5. Over what timeframe was the data collected? Does this timeframe match the creation timeframe of the data associated with the instances (e.g. recent crawl of old news articles)? If not, please describe the timeframe in which the data associated with the instances was created. Finally, list when the dataset was first published.

    The dataset was collected between May 2020 and July 2021.

  6. Were any ethical review processes conducted (e.g. by an institutional review board)? If so, please provide a description of these review processes, including the outcomes, as well as a link or other access point to any supporting documentation.

    No.

  7. Does the dataset relate to people? If not, you may skip the remainder of the questions in this section.

    The dataset is related to people in so far that the dataset creators are individual users of reddit and posts can contain people.

  8. Did you collect the data from the individuals in question directly, or obtain it via third parties or other sources (e.g. websites)?

    The dataset was collected via Reddit's API. Thus, only public posts and data was downloaded.

  9. Were the individuals in question notified about the data collection? If so, please describe (or show with screenshots or other information) how notice was provided, and provide a link or other access point to, or otherwise reproduce, the exact language of the notification itself.

    NA.

  10. Did the individuals in question consent to the collection and use of their data? If so, please describe (or show with screenshots or other information) how consent was requested and provided, and provide a link or other access point to, or otherwise reproduce, the exact language to which the individuals consented.

    NA.

  11. If consent was obtained, were the consenting individuals provided with a mechanism to revoke their consent in the future or for certain uses? If so, please provide a description, as well as a link or other access point to the mechanism (if appropriate).

    NA.

  12. Has an analysis of the potential impact of the dataset and its use on data subjects (e.g. a data protection impact analysis) been conducted? If so, please provide a description of this analysis, including the outcomes, as well as a link or other access point to any supporting documentation.

    NA.

  13. Any other comments?

    No.

Preprocessing / Cleaning / Labeling

Dataset creators should read through these questions prior to any pre-processing, cleaning, or labeling and then provide answers once these tasks are complete. The questions in this section are intended to provide dataset consumers with the information they need to determine whether the “raw” data has been processed in ways that are compatible with their chosen tasks. For example, text that has been converted into a “bag-of-words” is not suitable for tasks involving word order.

  1. Was any preprocessing/cleaning/labeling of the data done (e.g. discretization or bucketing, tokenization, part-of-speech tagging, SIFT feature extraction, removal of instances, processing of missing values)? If so, please provide a description. If not, you may skip the remainder of the questions in this section.

    The released dataset was preprocessed using an automated pipeline. This pipeline was taken from PASS and was used to removed videos that contain human faces using a publicly available face classifier.

  2. Was the "raw" data saved in addition to the preprocessed/cleaned/labeled data (e.g. to support unanticipated future uses)? If so, please provide a link or other access point to the "raw" data.

    Yes.

  3. Is the software used to preprocess/clean/label the instances available? If so, please provide a link or other access point.

    Yes. We used detoxify to filter unsafe text:

  4. Any other comments?

    No.

Uses

These questions are intended to encourage dataset creators to reflect on the tasks for which the dataset should and should not be used. By explicitly highlighting these tasks, dataset creators can help dataset consumers to make informed decisions, thereby avoiding potential risks or harms.

  1. Has the dataset been used for any tasks already? If so, please provide a description.

    This dataset has only been used for the experiments in this paper.

  2. Is there a repository that links to any or all papers or systems that use the dataset? If so, please provide a link or other access point.

    Google scholar will be able to track which papers have built upon this dataset/idea.

  3. What (other) tasks could the dataset be used for?

    This dataset can be used for multi-modal representation learning or video-text retrieval.

  4. Is there anything about the composition of the dataset or the way it was collected and preprocessed/cleaned/labeled that might impact future uses? For example, is there anything that a future user might need to know to avoid uses that could result in unfair treatment of individuals or groups (e.g. stereotyping, quality of service issues) or other undesirable harms (e.g. financial harms, legal risks) If so, please provide a description. Is there anything a future user could do to mitigate these undesirable harms?

    Not that we are aware of.

  5. Are there tasks for which the dataset should not be used? If so, please provide a description.

    This dataset should not be used for tasks that might disclose the identity of the users or directly or indirectly harm them.

  6. Any other comments?

    No.

Distribution

Dataset creators should provide answers to these questions prior to distributing the dataset either internally within the entity on behalf of which the dataset was created or externally to third parties.

  1. Will the dataset be distributed to third parties outside of the entity (e.g. company, institution, organization) on behalf of which the dataset was created? If so, please provide a description.

    No.

  2. How will the dataset will be distributed (e.g. tarball on website, API, GitHub)? Does the dataset have a digital object identifier (DOI)?

    The dataset will have a website and GitHub repository and be downloaded as a csv file containing links to the data points.

  3. When will the dataset be distributed?

    The dataset will be published together with this paper.

  4. Will the dataset be distributed under a copyright or other intellectual property (IP) license, and/or under applicable terms of use (ToU)? If so, please describe this license and/or ToU, and provide a link or other access point to, or otherwise reproduce, any relevant licensing terms or ToU, as well as any fees associated with these restrictions.

    The dataset will be distributed under a research license.

  5. Have any third parties imposed IP-based or other restrictions on the data associated with the instances? If so, please describe these restrictions, and provide a link or other access point to, or otherwise reproduce, any relevant licensing terms, as well as any fees associated with these restrictions.

    No.

  6. Do any export controls or other regulatory restrictions apply to the dataset or to individual instances? If so, please describe these restrictions, and provide a link or other access point to, or otherwise reproduce, any supporting documentation.

    NA.

  7. Any other comments?

    No.

Maintenance

As with the previous section, dataset creators should provide answers to these questions prior to distributing the dataset. These questions are intended to encourage dataset creators to plan for dataset maintenance and communicate this plan with dataset consumers.

  1. Who is supporting/hosting/maintaining the dataset?

    The authors will maintain the dataset. In particular, Laura Hanu, who can be contacted at laura@unitary.ai

  2. How can the owner/curator/manager of the dataset be contacted (e.g. email address)?

    The website of the dataset will contain all information to contact the authors and or maintainers of the dataset.

  3. Is there an erratum? If so, please provide a link or other access point.

    No.

  4. Will the dataset be updated (e.g. to correct labeling errors, add new instances, delete instances)? If so, please describe how often, by whom, and how updates will be communicated to users (e.g. mailing list, GitHub)?

    Yes, the website will contain a mechanism to version and update the dataset in case of errors.

  5. If the dataset relates to people, are there applicable limits on the retention of the data associated with the instances (e.g. were individuals in question told that their data would be retained for a fixed period of time and then deleted)? If so, please describe these limits and explain how they will be enforced.

    Not to our knowledge.

  6. Will older versions of the dataset continue to be supported/hosted/maintained? If so, please describe how. If not, please describe how its obsolescence will be communicated to users.

    Yes, through versioning on GitHub.

  7. If others want to extend/augment/build on/contribute to the dataset, is there a mechanism for them to do so? If so, please provide a description. Will these contributions be validated/verified? If so, please describe how. If not, why not? Is there a process for communicating/distributing these contributions to other users? If so, please provide a description.

    Yes, on the website.

  8. Any other comments?

    No.