Wikidata:WikiProject PersonalData/european elections

Goal edit

This subproject, initiated by PersonalDataIO, aims to develop a tool that can massively generate access requests to personal data-processing operations related to the European Elections of 2019.

Context edit

Although efforts towards re-gaining control over personal data are increasing, collaboration between technical-legal-activism arenas is still inefficient. We believe this is mainly due to lack of information on personal data flows through the ecosystem of actors (platforms, ad-tech companies, app trackers). This hinders not only informed decision making at the policy level, but also development of tools needed to exercise our data rights.

We're developing an open and collaborative methodology to tackle this problem by reconstructing the personal data ecosystem (mapping data controllers, which data they share, to whom) and creating tools to read that data and generate access requests. You can read more about the main project here Wikidata:WikiProject_PersonalData.

We think it's particularly interesting to have this mapping for the case of elections. We consider it can provide useful insights to understand how the profiling and distribution of targeting information works.

Why? edit

We think that by requesting your personal data from political parties, one can get more insights into the use of data by those parties, in particular for any profiling that might take place, and use on social media.

These techniques have been used during a few separate elections, to great success:

  • An individual (helped by Paul-Olivier Dehaye, leader of this project), asked for his data from Cambridge Analytica. In a first instance, this has led to substantial disclosure by the company of its data practices. In a second instance (ongoing), this has led to significant disclosure regardign various actors in the litigation proceedings.
  • Carole Cadwalladr (The Observer), helped by Dehaye, has requested her data from car insurance companies in the UK, which has led to the uncovering of significant data transfers between insurance companies and Leave.EU, one of the branches of the Leave campaign.

We don't expect the general public to spontaneously engage in this, but some combination of motivated citizens and journalists could use those techniques to create very effective transparency.

However we do expect reports to be based on such efforts, increasing global awareness of these rights, and increasing pressure on politicians to disclose more pro-actively how they use data in the context of elections.

Some of the questions we think are interesting around the map:

  • How is one voter’s personal data flowing?
  • Through which nodes?
  • How are these nodes inter-linked?
  • How is this configuration affecting the distribution of information?


Example in the 2019 context

For instance, as a particularly complex case: Manfred Weber (Q88866), Spitzenkandidat (Q60477938) for the European People's Party (Q208242), uses NationBuilder (Q29001016) for managing his campaign. Individuals registering to his mailing list have to agree "to be matched to [their] social media accounts so the EPP can send [them] information relevant to my interests", or at least he obtains consent for this. The privacy policy includes "This Notice applies to all personal data collected and used by the EPP about the general public. This Notice also applies to campaigns we carry out through third party social media sites (e.g. Facebook or Twitter), which we may use to gather public information about you". This all implies some level of profiling of the interested individuals. What other sources of data is Weber using? How extensive is the profiling done? Note that such practices have been under scrutiny in the past, and that there are very subtle differences in law on how to approach these questions from Member State to Member State.

How to edit

In order to run this project, collaboration from key stakeholders is required: journalists associations, academics, regulators, legal experts, technologists, data protection authorities, communities built around personal data empowerment.

We will use Wikidata to map nodes (or "data controllers" as specified by the General Data Protection Regulation), which data they are collecting, flows of data between them, presence/absence of privacy policies and more. In the context of European elections, some of the nodes will be national parties, supra national parties, individual candidates, and digital agencies they contract with. We refer to these activities as 'mapping' or 'writing'.

We will in parallel construct "thin" and privacy-preserving clients generating requests for access to one's data based on this mapping, which could be used by the general public or (first) journalists and motivated citizens. We refer to these activities as 'clients' or 'reading'.

Activities edit

These are the activities identified so far:

  1. Mapping elections, political parties, affiliations, politicians
    1. Map elections going on in parallel in 27 countries (status)
    2. Map political parties taking part in these elections (with the added complexity that they are actually organized through European parties/groups) (status)
    3. Map affiliation of national parties with supranational parties (for consistency mostly)
    4. Map politicians taking part in these elections (we possibly go with sufficient granularity to map some of them)
  1. Mapping the information necessary for making | Subject Access Requests
    1. Tagging these parties or politicians as likely holders of personal data
    2. URL for their privacy policy
    3. Contact email where to request your data

Mapping information on elections, political parties & politicians edit

The European elections are complex beasts. For instance, the parties actually allowed to run are only the supranational parties recognized in the official Register of European Political Parties (Q60310569). National parties affiliate themselves to supranational parties (or sometimes European Parliament groups?) to run at national level. Actually, in some countries, the electoral districts are *not* national, for instance Belgium or Poland[citation needed]

This is why we need to map not only elections but political parties at a national level, at a supranational level, their relationships and, if possible, politicians.

Listing elections themselves edit

Goal: List the elections in every country/district being run in the context of the 2019 European Parliament election (Q16999180)

In order to do this, we're going to build on the effort of the Wikidata:WikiProject_elections, a project that has already set a standard for how data on elections should be input in WikiData.

note: complete here with relevant info from Wikidata:WikiProject_elections to this project

Listing political parties edit

Goal: List all the parties participating in the European elections, either at national or supranational level

For example, here are some parties from Spain and Finland, involved in the European elections 2019:

note: look for template (or make one) for political parties

Mapping relations between supranational and national political parties edit

Goal: Identify relationships between political parties at the national and the supranational level - Who is a member of who?

The European elections are a bit special: supranational parties are only allowed to run. So if you access their list, you can see for each who are members, for instance by following: Party of European Socialists (Q220945)described at URL (P973) https://www.pes.eu/en/members/


  • The supranational political parties are listed here.
  • The national political parties are listed here.


note: define statements to use (look into Wikidata:WikiProject elections for reference

Mapping political parties and their relations to the elections edit

Goal: List all the parties participating in the European elections

This is still work in progress. You can take a look at this query: [1] for ongoing work.

note: check which statement Wikidata:WikiProject elections uses

Mapping politicians edit

Goal: List politicians participating in the European elections (at least the ones with interesting digital strategies)

In order to add politicians that do not appear in WikiData, we'll follow the data model proposed by Wikidata:WikiProject_every_politician.

note: provide relevant info here from Wikidata:WikiProject_every_politician note2: check Wikidata:WikiProject elections to see how they link politicians as participating in elections

Mapping information on personal data use edit

The second part of mapping means that we input information on which of these actors process personal data, which is their privacy policy and their contact e-mail. With this in place, access requests can be generated by the client tools.

Identifying data controllers edit

Goal: Identify all the parties (national and supranational) and politicians participating in the European elections, as data controller (Q59859818)

This can be done by adding the statement


Also, we need to link a data controller with its privacy policy. This can be done through the statement

Mapping privacy policies edit

Goal: generate items in WikiData for all data controller's privacy policies, with URL and contact e-mail

Items for data controller's privacy policies can be created following this model: PP (Spain) privacy policy (Q60178324))


The URL of the privacy policy should appear, via the statement

Mapping contact information edit

Goal: generate items in WikiData for all data controller's privacy policies, with URL and contact e-mail

You can input the e-mail of the data protection officer, through the statement

Tools for 'reading' the map edit

Client (to which we may also refer as 'reading') is the complementary side of the map. How do we obtain information from the map?

Using the data loaded into WikiData, we are building tools that can read it and produce visualizations, taking into account the questions mentioned above. For this, we're using Jupyter Notebooks and MyBinder.org.

How can I contribute? edit

It is important to highlight here that any step of this mapping process helps, as there is a wide variety of skills and knowledge involved in mapping (spoken language, computer language, political knowledge, wikidata knowledge, legal knowledge, "scaling up" knowledge, etc). So if you can contribute to any part of this process, please do.

As shown above, you can contribute by:

Listing elections

If you are aware of an election that is part of 2019 European Parliament election (Q16999180) and is not in this[2] list, you can create the item for the election and incorporate it to the project.

Please, first check the item is not already created in WikiData. Maybe it's there but it's not linked. If this is the case, just make sure the following statements are present on the item's page.

If you can't find it then you can create an item for an election. Again, make sure these statements are present in the item page:


And the inverse, please go to 2019 European Parliament election (Q16999180) and add the statement:


Mapping political parties and their relations to the elections

If the election for your country is mapped, you can check if the political parties participating of it are listed or not.

You can do this by checking on this list[3].

If the party doesn't appear in the list, first check if there is a WikiData page for it, maybe it's there but it's not linked. If this is the case, just make sure the following statements are present on the item's page.

If it's not in the list and it doesn't appear in WikiData, you can create it. Make sure the following statement is present:


Also, go to the local election page and update it with the statement:


Mapping political parties affiliation to supranational parties

You can access a list of supranational parties participating of the EU election here[4].

//work in progress

Participants edit

Who does what? edit

Pdehaye and Cassandreces are setting the initial infrastructure needed for others to collaborate. Pdehaye is also developing the client tools.

See also edit