ABOUT ML Reference Document

Last Updated

Appendix A: Compiled List of Documentation Questions

Appendix A: Compiled List of Documentation Questions

This is an incomplete list of documentation proposals/questions/formats from published academic papers. In the future, the ABOUT ML project may take on a multistakeholder process of surfacing common questions and creating a synthesized single list.

Fact Sheets (Arnold et al. 2018)

Fact Sheets (Arnold et al. 2018)

Source: Arnold, M., Bellamy, R.K.E., Hind, M., Houde, S., Mehta, S., Mojsilovic, A., Nair, R., Ramamurthy, K. N., Reimer, D. Olteanu, A., Piorkowski, D., Tsay, J., & Varshney, K. R. (2018). FactSheets: Increasing Trust in AI Services through Supplier’s Declarations of Conformity. arXiv preprint arXiv:1808.07261. https://arxiv.org/abs/1808.07261

A few examples of items a FactSheet might include are:

  • What is the intended use of the service output?
  • What algorithms or techniques does this service implement?
  • Which datasets was the service tested on? (Provide links to datasets that were used for testing, along with corresponding datasheets.)
  • Describe the testing methodology.
  • Describe the test results.
  • Are you aware of possible examples of bias, ethical issues, or other safety risks as a result of using the service?
  • Are the service outputs explainable and/or interpretable?
  • For each dataset used by the service: Was the dataset checked for bias? What efforts were made to ensure that it is fair and representative?
  • Does the service implement and perform any bias detection and remediation?
  • What is the expected performance on unseen data or data with different distributions?
  • Was the service checked for robustness against adversarial attacks?
  • When were the models last updated?

Data Sheets (Gebru et al. 2018)

Data Sheets (Gebru et al. 2018)

Source: Gebru, T., Morgenstern, J., Vecchione, B., Vaughan, J. W., Wallach, H., Daumeé III, H., & Crawford, K. (2018). Datasheets for datasets. arXiv preprint arXiv:1803.09010. https://arxiv.org/abs/1803.09010

Motivation
  • 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.
  • Who created this dataset (e.g., which team, research group) and on behalf of which entity (e.g., company, institution, organization)?
  • Who funded the creation of the dataset? If there is an associated grant, please provide the name of the grantor and the grant name and number.
  • Any other comments?
Composition
  • 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.
  • How many instances are there in total (of each type, if appropriate)?
  • 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).
  • What data does each instance consist of? “Raw” data (e.g., unprocessed text or images) or features? In either case, please provide a description.
  • Is there a label or target associated with each instance? If so, please provide a description.
  • 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, for example, redacted text.
  • 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.
  • 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.
  • Are there any errors, sources of noise, or redundancies in the dataset? If so, please provide a description.
  • 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.
  • 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.
  • Does the dataset contain data that, if viewed directly, might be offensive, insulting, threatening, or might otherwise cause anxiety? If so, please describe why.
  • Does the dataset relate to people? If not, you may skip the remaining questions in this section.
  • 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.
  • 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.
  • 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.
  • Any other comments?
Collection Process
  • 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.
  • 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?
  • If the dataset is a sample from a larger set, what was the sampling strategy (e.g., deterministic, probabilistic with specific sampling probabilities)?
  • 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)?
  • 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.
  • 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.
  • Does the dataset relate to people? If not, you may skip the remainder of the questions in this section.
  • Did you collect the data from the individuals in question directly, or obtain it via third parties or other sources (e.g., websites)?
  • 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.
  • 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.
  • 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).
  • 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.
  • Any other comments?
Preprocessing/Cleaning/Labeling
  • 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.
  • 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.
  • Is the software used to preprocess/clean/label the instances available? If so, please provide a link or other access point.
  • Any other comments?
Uses
  • Has the dataset been used for any tasks already? If so, please provide a description.
  • 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.
  • What (other) tasks could the dataset be used for?
  • 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?
  • Are there tasks for which the dataset should not be used? If so, please provide a description.
  • Any other comments?
Distribution
  • 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.
  • How will the dataset be distributed (e.g., tarball on website, API, GitHub)? Does the dataset have a digital object identifier (DOI)?
  • When will the dataset be distributed?
  • 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.
  • 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.
  • 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.
  • Any other comments?
Maintenance
  • Who is supporting/hosting/maintaining the dataset?
  • How can the owner/curator/manager of the dataset be contacted (e.g., email address)?
  • Is there an erratum? If so, please provide a link or other access point.
  • 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)?
  • 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.
  • 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.
  • 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.
  • Any other comments?

Model Cards (Mitchell et al. 2018)

Model Cards (Mitchell et al. 2018)

Source: Mitchell, M., Wu, S., Zaldivar, A., Barnes, P., Vasserman, L., Hutchinson, B., Spitzer E., Raji, I.D., & Gebru, T. (2019, January). Model cards for model reporting. In Proceedings of the Conference on Fairness, Accountability, and Transparency (pp. 220-229). ACM. https://arxiv.org/abs/1810.03993

Model Details

Basic information about the model.

  • Person or organization developing model
  • Model date
  • Model version
  • Model type
  • Information about training algorithms, parameters, fairness constraints or other applied approaches, and features
  • Paper or other resource for more information
  • Citation details
  • License
  • Where to send questions or comments about the model
Intended Use

Use cases that were envisioned during development.

  • Primary intended uses
  • Primary intended users
  • Out-of-scope use cases
Factors

Factors could include demographic or phenotypic groups, environmental conditions, technical attributes, or others listed in Section 4.3 of the Model Cards paper.

  •  Relevant factors
  •  Evaluation factors
Metrics

Metrics should be chosen to reflect potential real world impacts of the model.

  • Model performance measures
  • Decision thresholds
  • Variation approaches
Evaluation Data

Details on the dataset(s) used for the quantitative analyses in the card.

  • Datasets
  • Motivation
  • Preprocessing
Training Data.

May not be possible to provide in practice. When possible, this section should mirror Evaluation Data. If such detail is not possible, minimal allowable information should be provided here, such as details of the distribution over various factors in the training datasets.

Quantitative Analyses

  • Unitary results
  • Intersectional results

A “Nutrition Label” for Privacy (Kelley et al. 2009)

A “Nutrition Label” for Privacy (Kelley et al. 2009)

Source: Kelley, P. G., Bresee, J., Cranor, L. F., & Reeder, R. W. (2009, July). A nutrition label for privacy. In Proceedings of the 5th Symposium on Usable Privacy and Security (p. 4). ACM. http://cups.cs.cmu.edu/soups/2009/proceedings/a4-kelley.pdf

The Dataset Nutrition Label: A Framework To Drive Higher Data Quality Standards (Holland et al. 2019)

The Dataset Nutrition Label: A Framework To Drive Higher Data Quality Standards (Holland et al. 2019)

Source: Holland, S., Hosny, A., Newman, S., Joseph, J., & Chmielinski, K. (2018). The dataset nutrition label: A framework to drive higher data quality standards. arXiv preprint arXiv:1805.03677. https://arxiv.org/abs/1805.03677

  • What is the relevant metadata collected with the dataset?
  • Detail source, author contact information and version history.
  • Ground truth correlations: linear correlations between a chosen variable in the dataset and variables from other datasets considered to be “ground truth.”

Data Statements for Natural Language Processing: Toward Mitigating System Bias and Enabling Better Science (Bender and Friedman 2018)

Data Statements for Natural Language Processing: Toward Mitigating System Bias and Enabling Better Science (Bender and Friedman 2018)

Source: Bender, E. M., & Friedman, B. (2018). Data statements for natural language processing: Toward mitigating system bias and enabling better science. Transactions of the Association for Computational Linguistics, 6, 587-604. https://aclweb.org/anthology/papers/Q/Q18/Q18-1041/

Curation Rationale

Which texts were included and what were the goals in selecting texts, both in the original collection and in any further sub-selection? This can be especially important in datasets too large to thoroughly inspect by hand. An explicit statement of the curation rationale can help dataset users make inferences about what other kinds of texts systems trained with them could conceivably generalize to.

Language Variety

Languages differ from each other in structural ways that can interact with NLP algorithms. Within a language, regional or social dialects can also show great variation (Chambers and Trudgill, 1998).

The language and language variety should be described with:

  • A language tag from BCP-479 identifying the language variety (e.g. en-US or yue-HantHK)
  • A prose description of the language variety, glossing the BCP-47 tag and also providing further information (e.g. English as spoken in Palo Alto CA (USA) or Cantonese written with traditional characters by speakers in Hong Kong who are bilingual in Mandarin)
Speaker Demographic

Sociolinguistics has found that variation (in pronunciation, prosody, word choice, and grammar) correlates with speaker demographic characteristics (Labov, 1966), as speakers use linguistic variation to construct and project identities (Eckert and Rickford, 2001). Transfer from native languages (L1) can affect the language produced by non-native (L2) speakers (Ellis, 1994, Ch. 8). A further important type of variation is disordered speech (e.g. dysarthria).

Specifications include:

  • Age
  • Gender
  • Race/ethnicity
  • Native language
  • Socio-economic status
  • Number of different speakers represented
  • Presence of disordered speech
Annotator Demographic

What are the demographic characteristics of the annotators and annotation guideline developers? Their own ‘social address’ influences their experience with language and thus their perception of what they are annotating.

Specifications include:

  • Age
  • Gender
  • Race/ethnicity
  • Native language
  • Socio-economic status
  • Training in linguistics/other relevant discipline
Speech Situation

Characteristics of the speech situation can affect linguistic structure and patterns at many levels. The intended audience of a linguistic performance can also affect linguistic choices on the part of speakers. The time and place provide broader context for understanding how the texts collected relate to their historical moment and should also be made evident in the data statement.

Specifications include:

  • Time and place
  • Modality (spoken/signed, written)
  • Scripted/edited v. spontaneous
  • Synchronous v. asynchronous interaction
  • Intended audience
Text Characteristics

Both genre and topic influence the vocabulary and structural characteristics of texts (Biber, 1995), and should be specified.

Recording Quality

For data that includes audio/visual recordings, indicate the quality of the recording equipment and any aspects of the recording situation that could impact recording quality.

Other

There may be other information of relevance as well (e.g. the demographic characteristics of the curators). As stated above, this is intended as a starting point and we anticipate best practices around writing data statements to develop over time.

Provenance Appendix

For datasets built out of existing datasets, the data statements for the source datasets should be included as an appendix.

ABOUT ML Reference Document

Section 0: How to Use this Document

Recommended Reading Plan

Quick Guides

How We Define

Contact for Support

Section 1: Project Overview

1.1 Statement of Importance for ABOUT ML Project

1.1.0 Importance of Transparency: Why a Company Motivated by the Bottom Line Should Adopt ABOUT ML Recommendations

1.1.1 About This Document and Version Numbering

1.1.2 ABOUT ML Goals and Plan

1.1.3 ABOUT ML Project Process and Timeline Overview

1.1.4 Who Is This Project For?

1.1.4.1 Audiences for the ABOUT ML Resources

1.1.4.2 Stakeholders That Should Be Consulted While Putting Together ABOUT ML Resources

1.1.4.3 Audiences for ABOUT ML Documentation Artifacts

1.1.4.4 Whose Voices Are Currently Reflected in ABOUT ML?

1.1.4.5 Origin Story

Section 2: Literature Review (Current Recommendations on Documentation for Transparency in the ML Lifecycle)

2.1 Demand for Transparency and AI Ethics in ML Systems 

2.2 Documentation to Operationalize AI Ethics Goals

2.2.1 Documentation as a Process in the ML Lifecycle

2.2.2 Key Process Considerations for Documentation

2.3 Research Themes on Documentation for Transparency 

2.3.1 System Design and Set Up

2.3.2 System Development

2.3.3 System Deployment

Section 3: Preliminary Synthesized Documentation Suggestions

3.4.1 Suggested Documentation Sections for Datasets

3.4.1.1 Data Specification

3.4.1.1.1 Motivation

3.4.1.2 Data Curation 

3.4.1.2.1 Collection

3.4.1.2.2 Processing

3.4.1.2.3 Composition

3.4.1.2.4 Types and Sources of Judgement Calls

3.4.1.3 Data Integration

3.4.1.3.1 Use

3.4.1.3.2 Distribution

3.4.1.4 Maintenance

3.4.2 Suggested Documentation Sections for Models

3.4.2.1 Model Specifications

3.4.2.2 Model Training

3.4.2.3 Evaluation

3.4.2.4 Model Integration

3.4.2.5 Maintenance

Section 4: Current Challenges of Implementing Documentation

Section 5: Conclusions

Version 0

Version 1

Appendix A: Compiled List of Documentation Questions 

Fact Sheets (Arnold et al. 2018)

Data Sheets (Gebru et al. 2018)

Model Cards (Mitchell et al. 2018)

A “Nutrition Label” for Privacy (Kelley et al. 2009)

The Dataset Nutrition Label: A Framework To Drive Higher Data Quality Standards (Holland et al. 2019)

Data Statements for Natural Language Processing: Toward Mitigating System Bias and Enabling Better Science (Bender and Friedman 2018)

Appendix B: Diverse Voices Process and Artifacts

Procurement Recruitment Email

Procurement Confirmation Email 

Appendix C: Glossary

Sources Cited

  1. Holstein, K., Vaughan, J.W., Daumé, H., Dudík, M., u0026amp; Wallach, H.M. (2018). Improving Fairness in Machine Learning Systems: What Do Industry Practitioners Need? CHI.
  2. Young, M., Magassa, L. and Friedman, B. (2019) Toward inclusive tech policy design: a method for underrepresented voices to strengthen tech policy documents. Ethics and Information Technology 21(2), 89-103.
  3. World Wide Web Consortium Process Document (W3C) process outlined here: https://www.w3.org/2019/Process-20190301/
  4. Internet Engineering Task Force (IETF) process outlined here: https://www.ietf.org/standards/process/
  5. The Web Hypertext Application Technology Working Group (WHATWG) process outlined here: https://whatwg.org/faq#process
  6. Oever, N., Moriarty, K. The Tao of IETF: A novice's guide to the Internet Engineering Task Force. https://www.ietf.org/about/participate/tao/.
  7. Young, M., Magassa, L. and Friedman, B. (2019) Toward inclusive tech policy design: a method for underrepresented voices to strengthen tech policy documents. Ethics and Information Technology 21(2), 89-103.
  8. Friedman, B, Kahn, Peter H., and Borning, A., (2008) Value sensitive design and information systems. In Kenneth Einar Himma and Herman T. Tavani (Eds.) The Handbook of Information and Computer Ethics., (pp. 70-100) John Wiley u0026amp; Sons, Inc. http://jgustilo.pbworks.com/f/the-handbook-of-information-and-computer-ethics.pdf#page=104; Davis, J., and P. Nathan, L. (2015). Value sensitive design: applications, adaptations, and critiques. Handbook of Ethics, Values, and Technological Design: Sources, Theory, Values and Application Domains. (pp. 11-40) DOI: 10.1007/978-94-007-6970-0_3. https://www.researchgate.net/publication/283744306_Value_Sensitive_Design_Applications_Adaptations_and_Critiques; Borning, A. and Muller, M. (2012). Next steps for value sensitive design. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI '12). (pp 1125-1134) DOI: https://doi.org/10.1145/2207676.2208560 https://dl.acm.org/citation.cfm?id=2208560
  9. Pichai, S., (2018). AI at Google: our principles. The Keyword. https://www.blog.google/technology/ai/ai-principles/; IBM’s Principles for Trust and Transparency. IBM Policy. https://www.ibm.com/blogs/policy/trust-principles/; Microsoft AI principles. Microsoft. https://www.microsoft.com/en-us/ai/our-approach-to-ai; Ethically Aligned Design – Version II. IEEE. https://standards.ieee.org/content/dam/ieee-standards/standards/web/documents/other/ead_v2.pdf
  10. Zeng, Y., Lu, E., and Huangfu, C. (2018) Linking artificial intelligence principles. CoRR https://arxiv.org/abs/1812.04814.
  11. essica Fjeld, Hannah Hilligoss, Nele Achten, Maia Levy Daniel, Sally Kagay, and Joshua Feldman, (2018). Principled artificial intelligence - a map of ethical and rights based approaches, Berkman Center for Internet and Society, https://ai-hr.cyber.harvard.edu/primp-viz.html
  12. Jobin, A., Ienca, M., u0026amp; Vayena, E. (2019). Artificial Intelligence: the global landscape of ethics guidelines. arXiv preprint arXiv:1906.11668. https://arxiv.org/pdf/1906.11668.pdf
  13. Jobin, A., Ienca, M., u0026amp; Vayena, E. (2019). Artificial Intelligence: the global landscape of ethics guidelines. arXiv preprint arXiv:1906.11668. https://arxiv.org/pdf/1906.11668.pdf
  14. Ananny, M., and Kate Crawford (2018). Seeing without knowing: Limitations of the transparency ideal and its application to algorithmic accountability. New Media and Society 20 (3): 973-989.
  15. Whittlestone, J., Nyrup, R., Alexandrova, A., u0026amp; Cave, S. (2019, January). The Role and Limits of Principles in AI Ethics: Towards a Focus on Tensions. In Proceedings of the AAAI/ACM Conference on AI Ethics and Society, Honolulu, HI, USA (pp. 27-28). http://www.aies-conference.com/wp-content/papers/main/AIES-19_paper_188.pdf; Mittelstadt, B. (2019). AI Ethics–Too Principled to Fail? https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3391293
  16. Greene, D., Hoffmann, A. L., u0026amp; Stark, L. (2019, January). Better, nicer, clearer, fairer: A critical assessment of the movement for ethical artificial intelligence and machine learning. In Proceedings of the 52nd Hawaii International Conference on System Sciences. https://scholarspace.manoa.hawaii.edu/handle/10125/59651
  17. Raji, I. D., u0026amp; Buolamwini, J. (2019). Actionable auditing: Investigating the impact of publicly naming biased performance results of commercial ai products. In AAAI/ACM Conf. on AI Ethics and Society (Vol. 1). https://www.media.mit.edu/publications/actionable-auditing-investigating-the-impact-of-publicly-naming-biased-performance-results-of-commercial-ai-products/
  18. Algorithmic Impact Assessment (2019) Government of Canada https://www.canada.ca/en/government/system/digital-government/modern-emerging-technologies/responsible-use-ai/algorithmic-impact-assessment.html
  19. Benjamin, M., Gagnon, P., Rostamzadeh, N., Pal, C., Bengio, Y., u0026amp; Shee, A. (2019). Towards Standardization of Data Licenses: The Montreal Data License. arXiv preprint arXiv:1903.12262. https://arxiv.org/abs/1903.12262; Responsible AI Licenses v0.1. RAIL: Responsible AI Licenses. https://www.licenses.ai/ai-licenses
  20. See Citation 5
  21. Safe Face Pledge. https://www.safefacepledge.org/; Montreal Declaration on Responsible AI. Universite de Montreal. https://www.montrealdeclaration-responsibleai.com/; The Toronto Declaration: Protecting the right to equality and non-discrimination in machine learning systems. (2018). Amnesty International and Access Now. https://www.accessnow.org/cms/assets/uploads/2018/08/The-Toronto-Declaration_ENG_08-2018.pdf ; Dagsthul Declaration on the application of machine learning and artificial intelligence for social good. https://www.dagstuhl.de/fileadmin/redaktion/Programm/Seminar/19082/Declaration/Declaration.pdf
  22. Dobbe, R., Dean, S., Gilbert, T., u0026amp; Kohli, N. (2018). A Broader View on Bias in Automated Decision-Making: Reflecting on Epistemology and Dynamics. https://arxiv.org/pdf/1807.00553.pdf
  23. Wagstaff, K. (2012). Machine learning that matters. https://arxiv.org/pdf/1206.4656.pdf ; Friedman, B., Kahn, P. H., Borning, A., u0026amp; Huldtgren, A. (2013). Value sensitive design and information systems. In Early engagement and new technologies: Opening up the laboratory (pp. 55-95). Springer, Dordrecht. https://vsdesign.org/publications/pdf/non-scan-vsd-and-information-systems.pdf
  24. Dobbe, R., Dean, S., Gilbert, T., u0026amp; Kohli, N. (2018). A Broader View on Bias in Automated Decision-Making: Reflecting on Epistemology and Dynamics. https://arxiv.org/pdf/1807.00553.pdf
  25. Safe Face Pledge. https://www.safefacepledge.org/
  26. Montreal Declaration on Responsible AI. Universite de Montreal. https://www.montrealdeclaration-responsibleai.com/
  27. Diverse Voices How To Guide. Tech Policy Lab, University of Washington. https://techpolicylab.uw.edu/project/diverse-voices/
  28. Bender, E. M., u0026amp; Friedman, B. (2018). Data statements for natural language processing: Toward mitigating system bias and enabling better science. Transactions of the Association for Computational Linguistics, 6, 587-604.
  29. Ethically Aligned Design – Version II. IEEE. https://standards.ieee.org/content/dam/ieee-standards/standards/web/documents/other/ead_v2.pdf
  30. Gebru, T., Morgenstern, J., Vecchione, B., Vaughan, J. W., Wallach, H., Daumeé III, H., u0026amp; Crawford, K. (2018). Datasheets for datasets. https://arxiv.org/abs/1803.09010 https://arxiv.org/abs/1803.09010; Hazard Communication Standard: Safety Data Sheets. Occupational Safety and Health Administration, US Department of Labor. https://www.osha.gov/Publications/OSHA3514.html
  31. Holland, S., Hosny, A., Newman, S., Joseph, J., u0026amp; Chmielinski, K. (2018). The dataset nutrition label: A framework to drive higher data quality standards. https://arxiv.org/abs/1805.03677; Kelley, P. G., Bresee, J., Cranor, L. F., u0026amp; Reeder, R. W. (2009). A nutrition label for privacy. In Proceedings of the 5th Symposium on Usable Privacy and Security (p. 4). ACM. http://cups.cs.cmu.edu/soups/2009/proceedings/a4-kelley.pdf
  32. Mitchell, M., Wu, S., Zaldivar, A., Barnes, P., Vasserman, L., Hutchinson, B., ... u0026amp; Gebru, T. (2019, January). Model cards for model reporting. In Proceedings of the Conference on Fairness, Accountability, and Transparency (pp. 220-229). ACM. https://arxiv.org/abs/1810.03993
  33. Hind, M., Mehta, S., Mojsilovic, A., Nair, R., Ramamurthy, K. N., Olteanu, A., u0026amp; Varshney, K. R. (2018). Increasing Trust in AI Services through Supplier's Declarations of Conformity. https://arxiv.org/abs/1808.07261
  34. Veale M., Van Kleek M., u0026amp; Binns R. (2018) ‘Fairness and Accountability Design Needs for Algorithmic Support in High-Stakes Public Sector Decision-Making’ in Proceedings of the ACM Conference on Human Factors in Computing Systems, CHI 2018. https://arxiv.org/abs/1802.01029.
  35. Benjamin, M., Gagnon, P., Rostamzadeh, N., Pal, C., Bengio, Y., u0026amp; Shee, A. (2019). Towards Standardization of Data Licenses: The Montreal Data License. https://arxiv.org/abs/1903.12262
  36. Cooper, D. M. (2013, April). A Licensing Approach to Regulation of Open Robotics. In Paper for presentation for We Robot: Getting down to business conference, Stanford Law School.
  37. Responsible AI Practices. Google AI. https://ai.google/education/responsible-ai-practices
  38. Everyday Ethics for Artificial Intelligence. (2019). IBM. https://www.ibm.com/watson/assets/duo/pdf/everydayethics.pdf
  39. Federal Trade Commission. (2012). Best Practices for Common Uses of Facial Recognition Technologies (Staff Report). Federal Trade Commission, 30. https://www.ftc.gov/sites/default/files/documents/reports/facing-facts-best-practices-common-uses-facial-recognition-technologies/121022facialtechrpt.pdf
  40. Microsoft (2018). Responsible bots: 10 guidelines for developers of conversational AI. https://www.microsoft.com/en-us/research/uploads/prod/2018/11/Bot_Guidelines_Nov_2018.pdf
  41. Tramer, F., Atlidakis, V., Geambasu, R., Hsu, D., Hubaux, J. P., Humbert, M., ... u0026amp; Lin, H. (2017, April). FairTest: Discovering unwarranted associations in data-driven applications. In 2017 IEEE European Symposium on Security and Privacy (EuroSu0026amp;P) (pp. 401-416). IEEE. https://github.com/columbia/fairtest, https://www.mhumbert.com/publications/eurosp17.pdf
  42. Kishore Durg (2018). Testing AI: Teach and Test to raise responsible AI. Accenture Technology Blog. https://www.accenture.com/us-en/insights/technology/testing-AI
  43. Kush R. Varshney (2018). Introducing AI Fairness 360. IBM Research Blog. https://www.ibm.com/blogs/research/2018/09/ai-fairness-360/
  44. Dave Gershgorn (2018). Facebook says it has a tool to detect bias in its artificial intelligence. Quartz. https://qz.com/1268520/facebook-says-it-has-a-tool-to-detect-bias-in-its-artificial-intelligence/
  45. James Wexler. (2018) The What-If Tool: Code-Free Probing of Machine Learning Models. Google AI Blog. https://ai.googleblog.com/2018/09/the-what-if-tool-code-free-probing-of.html
  46. Miro Dudík, John Langford, Hanna Wallach, and Alekh Agarwal (2018). Machine Learning for fair decisions. Microsoft Research Blog. https://www.microsoft.com/en-us/research/blog/machine-learning-for-fair-decisions/
  47. Veale, M., Binns, R., u0026amp; Edwards, L. (2018). Algorithms that Remember: Model Inversion Attacks and Data Protection Law. Phil. Trans. R. Soc. A, 376, 20180083. https://doi.org/10/gfc63m
  48. Floridi, L. (2010, February). Information: A Very Short Introduction.
  49. Data Information Specialists Committee UK, 2007. http://www.disc-uk.org/qanda.html.
  50. Harwell, Drew. “Federal Study Confirms Racial Bias of Many Facial-Recognition Systems, Casts Doubt on Their Expanding Use.” The Washington Post, WP Company, 21 Dec. 2019, www.washingtonpost.com/technology/2019/12/19/federal-study-confirms-racial-bias-many-facial-recognition-systems-casts-doubt-their-expanding-use/
  51. Hildebrandt, M. (2019) ‘Privacy as Protection of the Incomputable Self: From Agnostic to Agonistic Machine Learning’, Theoretical Inquiries in Law, 20(1) 83–121.
  52. D'Amour, A., Heller, K., Moldovan, D., Adlam, B., Alipanahi, B., Beutel, A., ... u0026amp; Sculley, D. (2020). Underspecification presents challenges for credibility in modern machine learning. arXiv preprint arXiv:2011.03395.
  53. Selinger, E. (2019). ‘Why You Can’t Really Consent to Facebook’s Facial Recognition’, One Zero. https://onezero.medium.com/why-you-cant-really-consent-to-facebook-s-facial-recognition-6bb94ea1dc8f
  54. Lum, K., u0026amp; Isaac, W. (2016). To predict and serve?. Significance, 13(5), 14-19. https://rss.onlinelibrary.wiley.com/doi/full/10.1111/j.1740-9713.2016.00960.x
  55. LabelInsight (2016). “Drive Long-Term Trust u0026amp; Loyalty Through Transparency”. https://www.labelinsight.com/Transparency-ROI-Study
  56. Crawford and Paglen, https://www.excavating.ai/
  57. Geva, Mor u0026amp; Goldberg, Yoav u0026amp; Berant, Jonathan. (2019). Are We Modeling the Task or the Annotator? An Investigation of Annotator Bias in Natural Language Understanding Datasets. https://arxiv.org/pdf/1908.07898.pdf
  58. Bender, E. M., u0026amp; Friedman, B. (2018). Data statements for natural language processing: Toward mitigating system bias and enabling better science. Transactions of the Association for Computational Linguistics, 6, 587-604.
  59. Desmond U. Patton et al (2017).
  60. See Cynthia Dwork et al.,
  61. Katta Spiel, Oliver L. Haimson, and Danielle Lottridge. (2019). How to do better with gender on surveys: a guide for HCI researchers. Interactions. 26, 4 (June 2019), 62-65. DOI: https://doi.org/10.1145/3338283
  62. A. Doan, A. Y. Halevy, and Z. G. Ives. Principles of Data Integration. Morgan Kaufmann, 2012
  63. Momin M. Malik. (2019). Can algorithms themselves be biased? Medium. https://medium.com/berkman-klein-center/can-algorithms-themselves-be-biased-cffecbf2302c
  64. Fire, Michael, and Carlos Guestrin (2019). “Over-Optimization of Academic Publishing Metrics: Observing Goodhart’s Law in Action.” GigaScience 8 (giz053). https://doi.org/10.1093/gigascience/giz053.
  65. Vogelsang, A., u0026amp; Borg, M. (2019, September). Requirements engineering for machine learning: Perspectives from data scientists. In 2019 IEEE 27th International Requirements Engineering Conference Workshops (REW) (pp. 245-251). IEEE
  66. Eckersley, P. (2018). Impossibility and Uncertainty Theorems in AI Value Alignment (or why your AGI should not have a utility function). arXiv preprint arXiv:1901.00064.
  67. Partnership on AI. Report on Algorithmic Risk Assessment Tools in the U.S. Criminal Justice System, Requirement 5.
  68. Eckersley, P. (2018). Impossibility and Uncertainty Theorems in AI Value Alignment (or why your AGI should not have a utility function). arXiv preprint arXiv:1901.00064.https://arxiv.org/abs/1901.00064
  69. If it is not, there is likely a bug in the code. Checking a predictive model's performance on the training set cannot distinguish irreducible error (which comes from intrinsic variance of the system) from error introduced by bias and variance in the estimator; this is universal, and has nothing to do with different settings or
  70. Selbst, Andrew D. and Boyd, Danah and Friedler, Sorelle and Venkatasubramanian, Suresh and Vertesi, Janet (2018). “Fairness and Abstraction in Sociotechnical Systems”, ACM Conference on Fairness, Accountability, and Transparency (FAT*). https://ssrn.com/abstract=3265913
  71. Tools that can be used to explore and audit the predictive model fairness include FairML, Lime, IBM AI Fairness 360, SHAP, Google What-If Tool, and many others
  72. Wagstaff, K. (2012). Machine learning that matters. arXiv preprint arXiv:1206.4656. https://arxiv.org/abs/1206.4656
Table of Contents
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16