Dutch Algorithmic Transparency Standard

Mapping

✔️ UK:name matches NL:name

✔️ UK:description matches NL:description

✔️ UK:website matches NL:website

✔️ UK:contact_email matches NL:contact_email

🔔 UK:organisation matches NL:organization (notice: UK versus US spelling)

🔔 UK:team matches NL:department (notice: not sure if this is identical)

🔔 UK:senior_responsible_owner matches NL:competent_authority (notice: NL is a bit less personal here)

⚠️ UK:external_supplier* matches NL:application_url (notice: not quite the same thing)

⛔ UK:external_supplier_role* matches NL:NO MATCH

⛔ UK:external_supplier_identifier* matches NL:NO MATCH

⛔ UK:data_access_terms* matches NL:NO MATCH (notice: combine with NL:source_data?)

⚠️ UK:scope matches NL:goal (notice: NL:goal combines UK:scope and UK:benefit, also maybe overlaps with NL:proportionality)

⚠️ UK:benefit matches NL:goal (notice: NL:goal combines UK:scope and UK:benefit, also maybe overlaps with NL:proportionality)

⛔ UK:alternatives_considered matches NL:NO MATCH (notice: subsidiarity?)

Not yet added to the mapping

UK properties

The attribute 'model_type' indicates which types of methods or models the algorithm is using. For example expert system, deep neural network and so on.

- UK: model_type
  NL: 
  remark:
        

The attribute 'frequency_usage' gives information on how regularly the algorithmic tool is being used. For example the number of decisions made per month, the number of citizens interacting with the tool, and so on.

- UK: frequency_usage
  NL: 
  remark:
        

The attribute 'phase' indicates in which of the following stages or phases the tool is currently situated: - idea - design - development - production - retired This field includes date and time stamps of creation and any updates.

- UK: phase
  NL: 
  remark:
        

The attribute 'maintenance' gives details on the maintenance schedule and frequency of any reviews. For example, specific details on when and how a person reviews or checks the automated decision.

- UK: maintenance
  NL: 
  remark:
        

The attribute 'system_architecture' is the URL reference to documentation about the system architecture. For example, a link to a GitHub repository image or additional documentation about the system architecture.

- UK: system_architecture
  NL: 
  remark:
        

The attribute 'process_integration' explains how the algorithmic tool is integrated into the decision-making process and what influence the algorithmic tool has on the decision-making process. It gives a more detailed and extensive description of the wider decision-making process into which the algorithmic tool is embedded.

- UK: process_integration
  NL: 
  remark:
        

The attribute 'provided information” details how much and what information the algorithmic tool provides to the decision maker.

- UK: provided_information
  NL: 
  remark:
        

The attribute 'human_decisions' describes the decisions that people take in the overall process, including human review options.

- UK: human_decisions
  NL: 
  remark:
        

The attribute 'required_training' details the required training those deploying or using the algorithmic tool must undertake, if applicable; For example, the person responsible for the management of the tool had to complete data science training.

- UK: required_training
  NL: 
  remark:
        

The attribute 'appeals_and_review' details the mechanisms that are in place for review or appeal of the decision available to the general public.

- UK: appeals_and_review
  NL: 
  remark:
        

The attribute 'source_data_name' gives, if applicable, the name of the datasets used.

- UK: source_data_name*
  NL: 
  remark:
        

The attribute 'source_data' gives an overview of the data used to train and run the algorithmic tool. It will also specify whether data is used for training, testing, or operating. It should include which categories of data - for example 'age' or 'address' - which were used to train the model and which are used as input data for making a prediction.

- UK: source_data_description*
  NL: 
  remark:
        

The attribute 'source_data_url' provides a URL to the dataset wherever possible.

- UK: source_data_url*
  NL: 
  remark:
        

The attribute 'data_collection' gives information on the data collection process, including the original purpose of data collection.

- UK: data_collection*
  NL: 
  remark:
        

The attribute 'data_sharing_agreements' provides further information on data sharing agreements in place.

- UK: data_sharing_agreements*
  NL: 
  remark:
        

The attribute 'data_access_and_storage' details on who has or will have access to this data, how long it's stored, under what circumstances and by whom.

- UK: data_access_and_storage*
  NL: 
  remark:
        

The attribute 'impact_assessment_name' is a name and a short overview of the impact assessment conducted.

- UK: impact_assessment_name*
  NL: 
  remark:
        

The attribute 'impact_assessment_description' is a description of the impact assessments conducted.

- UK: impact_assessment_description*
  NL: 
  remark:
        

The attribute 'impact_assessment_date' is a date of when the impact assessment was conducted.

- UK: impact_assessment_date*
  NL: 
  remark:
        

The attribute 'impact_assessment_link' is a link to the impact assessment.

- UK: impact_assessment_link*
  NL: 
  remark:
        

The attribute 'risk_name' is an overview of the common risks for the algorithmic tool.

- UK: risk_name*
  NL: 
  remark:
        

The attribute 'risk_description' is a description of the risks identified under 'risk_name'.

- UK: risk_description*
  NL: 
  remark:
        

The attribute 'risk mitigation' is an overview of how the risks have been mitigated.

- UK: risk_mitigation*
  NL: 
  remark:
        

NL properties

Please give a short description of maximum 150 chars in order to give a quick overview of the purpose of the model, algorithm or AI. Example: 'The traffic light priority algorithm prioritises traffic modalities based on applicable law and local regulations'.

- NL: description_short
  UK: 
  remark:
        

Please indicate wether the model, algorithm or AI is descriptive, diagnostic, predictive or prescriptive. This information can be used as a possible risk indicator.

- NL: type
  UK: 
  remark:
        

Please provide keywords to facilitate search. Examples: 'traffic', 'transport', 'social security', 'crowd-monitoring', 'facial recognition', 'camera surveillance'.

- NL: category
  UK: 
  remark:
        

Please indicate wether the model, algorithm or AI is in development, in use, or archived.

- NL: status
  UK: 
  remark:
        

Please describe in what way citizens come into contact with the effects of the model, algorithm or AI. It should become clear under what specific circumstances this happens and what the expected consequences are on an individual and/or societal level.

- NL: impact
  UK: 
  remark:
        

Please describe why the authority by which the model, algorithm or AI is used is reasonably necessary. It should explain why the expected benefit outweighs any potential expected harm.

- NL: proportionality
  UK: 
  remark:
        

What is the official process in the organisation in which the model, algorithm or AI is involved. It should refer to concrete laws, regulation or policy, as published in publicly available sources.

- NL: decision_making_process
  UK: 
  remark:
        

Please provide a URL reference to any extended information about the use of the model, algorithm or AI within this specific use case.

- NL: documentation
  UK: 
  remark:
        

What is the URL reference to the the PublicCode.yml standard if available?

- NL: publiccode
  UK: 
  remark:
        

Please indicate wether a connection is being made to the Municipal Personal Records Database.

- NL: MPRD
  UK: 
  remark:
        

Please give an overview of the data that is being processed by the model, algorithm or AI. It should describe the pupose with which each data source is added and possible dependencies that resulting from this.

- NL: source_data
  UK: 
  remark:
        

Please indicate which standard methods or models the algorithm is using. Examples: ROC-curve, confusion matrix.

- NL: methods_and_models
  UK: 
  remark:
        

Please give a general overview of how the competent authority monitors the implementation of the model, algorithm or AI.

- NL: monitoring
  UK: 
  remark:
        

Please describe how the outcomes of the model, algorithm or AI can be intervened by humans. It should detail how the responsibility for possible human intervention is secured, so it's clear who can and may act.

- NL: human_intervention
  UK: 
  remark:
        

Please provide an overview of the outcome of the internal risk analysis. It can also refer to available online documentation. We currently refer to the assessment framework for algorithms by the Netherlands Court of Audit and the Regulation on a European Approach for Artificial Intelligence.

- NL: risks
  UK: 
  remark:
        

Please describe what the expected performance of the model, algorithm or AI is and how it is measured. It should detail which criteria are used and the frequency with which the performance is monitored.

- NL: performance_standard
  UK: 
  remark:
        

Please provide a link to the administrative act that makes the use of the model, algorithm or AI legitimate.

- NL: lawful_basis
  UK: 
  remark:
        

Has a data protection impact assessment been carried out?

- NL: DPIA
  UK: 
  remark:
        

Please give an overview of the key points from the data protection impact assessment. It should explain how discrimination is prevented when (proxies of) ethnicity, sex or zipcode are being used. If available it can reference the URL to the full DPIA documentation.

- NL: DPIA_description
  UK: 
  remark:
        

Please describe in what way can citizens object against the use or outcome of the model, algorithm or AI.

- NL: objection_procedure
  UK: 
  remark:
        

This is the schema used for this entry.

- NL: schema
  UK: 
  remark:
        

This is the Universal Unique Identifier for this entry.

- NL: id
  UK: 
  remark:
        

This is the URL for this entry.

- NL: url
  UK: 
  remark:
        

This is the geographical area to which this entry applies.

- NL: area
  UK: 
  remark:
        

This is the language in which this entry was filled.

- NL: lang
  UK: 
  remark:
        

This is the date before which this entry has to be revisited.

- NL: revision_date
  UK: 
  remark: