Gov/en/Portal:R&D/Innovations:Structured Data: Difference between revisions

(5 intermediate revisions by 2 users not shown)

Line 1:

{{KidsIntro|"Structured data" means organising information so computers can read it and connect it, like putting a clear label on every box so the right machine can find what is inside. On WikiDeal, some data is private (your identity, your bank details, photos of a flat you want to rent out) and some data is public on purpose (the content of contracts people choose to publish, whether a contract was signed or completed, and overall usage statistics). One idea guides the whole effort: your habits and your commercial behaviour are not for sale. It is a bit like nobody being allowed to walk into your head or your home: what you buy, how you behave, and your private practices stay yours. To picture it, think of a babysitter who keeps their own qualifications and references and can show them to a family when asked, without any platform storing or reselling them.}}

~~{{NotApproved}}~~

== Structured Data (Linked Data) ==

Line 19:

Line 17:

Structured Data is explored through several connected aspects, each a section below:

* [[#Why it matters|Why it matters]] (your habits are not for sale)

* [[#Why it matters|Why it matters]] (your habits are not for sale, user is not a product)

* [[#What data are we talking about|What data are we talking about]] (private vs published)

* [[#Three intended guarantees|Three intended guarantees]] (private by default, per-transaction control, trust circles)

Line 26:

Line 24:

* [[#Decentralised identity|Decentralised identity]] and the [[#Decentralized Web|Decentralized Web]]

* [[#Selective disclosure|Selective disclosure]]

* [[#Legal compliance: GDPR, eIDAS 2.0 and beyond|Legal compliance (GDPR, eIDAS 2.0)]]

* [[#Data governance: future hypotheses|Data governance: future hypotheses]]

=== Why it matters ===

On most platforms today, if a service is free, the user tends to be the product: behaviour is collected, analysed, and resold. WikiDeal '''is guided by''' the opposite principle. The intention is that there is no resale of users' habits or commercial tendencies, because the data is meant to belong to the users themselves.

On most platforms today, if a service is free of charge, the user tends to be the product: behaviour is collected, analysed, and resold (a pattern often described as [https://en.wikipedia.org/wiki/Surveillance_capitalism surveillance capitalism]). WikiDeal '''is guided by''' the opposite principle. The intention is that there is no resale of users' habits or commercial tendencies, because the data is meant to belong to the users themselves.

Put plainly, the goal is that nobody, no other [[Gov/en/Portal:R&D/Innovations:User Groups|User Group]], not the Ynternet.org Foundation, not the WikiDeal operator, gets to walk into your head or your home. Your buying patterns and private commercial practices are meant to stay yours.

Line 44:

Line 43:

'''Published data.''' Some data is meant to be public on WikiDeal: the content of contracts the parties choose to publish, the state of a contract (signed, completed, or ended in a dispute), and aggregate usage statistics. This data is published on purpose, not collected and resold.

The structured-data work '''intends to''' cover both kinds (expressing them as ~~'''~~RDF~~'''~~, ~~'''~~JSON-LD~~'''~~, and ~~'''~~Wikidata~~'''~~-linked resources) so each piece of information carries the right status and is machine-readable and interoperable. Governance of all this '''is intended to''' be decided in a '''participative context''', not dictated from above.

The structured-data work '''intends to''' cover both kinds (expressing them as [https://en.wikipedia.org/wiki/Resource_Description_Framework RDF], [https://en.wikipedia.org/wiki/JSON-LD JSON-LD], and [https://en.wikipedia.org/wiki/Wikidata Wikidata]-linked resources, in the spirit of [https://en.wikipedia.org/wiki/Linked_data Linked Data] and the [https://en.wikipedia.org/wiki/Semantic_Web Semantic Web]) so each piece of information carries the right status and is machine-readable and interoperable. Governance of all this '''is intended to''' be decided in a '''participative context''', not dictated from above.

=== Three intended guarantees ===

Line 73:

Line 72:

WikiDeal '''is designed to be guided by''' minimising the platform's access to user data. Where mainstream platforms treat user data as a primary asset, collecting, analysing, and monetising it, WikiDeal '''intends to''' treat data as belonging to users, held only in trust, for minimum periods, in minimum quantities.

The architecture WikiDeal '''is exploring''' is local-first: wherever possible, data would be stored on users' own devices, and only the minimum necessary information would be transmitted to shared Infrastructure. As an initial hypothesis, a babysitting agreement signed by both parties would live on the devices of the family and the babysitter, while the platform Infrastructure would hold only a cryptographic proof that the agreement was signed, not its content, unless both parties consent to shared storage.

The architecture WikiDeal '''is exploring''' is [https://en.wikipedia.org/wiki/Local-first_software local-first]: wherever possible, data would be stored on users' own devices, and only the minimum necessary information would be transmitted to shared Infrastructure. As an initial hypothesis, a babysitting agreement signed by both parties would live on the devices of the family and the babysitter, while the platform Infrastructure would hold only a cryptographic proof that the agreement was signed, not its content, unless both parties consent to shared storage.

This direction is technically challenging (it requires careful design of conflict resolution, backup, and recovery) but it '''aims at''' changing the data power relationship: a platform cannot sell data it does not have. The intention is that, even when WikiDeal eventually serves many babysitters, tutors, and freelancers, it would not have built a surveillance corpus of intimate human behaviour.

Line 79:

Line 78:

=== Decentralised identity ===

WikiDeal '''is evaluating''' decentralised identity approaches (DIDs, verifiable credentials) that '''aim at''' enabling trust between strangers without a central identity database. As above, a babysitter's qualifications and reputation could be held by the babysitter, presented on demand, and verified without the platform storing the underlying data. This research is ongoing through the WikiDeal [[Gov/en/Portal:R&D/Living-Labs|Living Labs]] programme.

WikiDeal '''is evaluating''' decentralised identity approaches ([https://en.wikipedia.org/wiki/Decentralized_identifier decentralised identifiers, DIDs], [https://en.wikipedia.org/wiki/Verifiable_credentials verifiable credentials], part of the wider [https://en.wikipedia.org/wiki/Self-sovereign_identity self-sovereign identity] field) that '''aim at''' enabling trust between strangers without a central identity database. As above, a babysitter's qualifications and reputation could be held by the babysitter, presented on demand, and verified without the platform storing the underlying data. This research is ongoing through the WikiDeal [[Gov/en/Portal:R&D/Living-Labs|Living Labs]] programme.

=== Decentralized Web ===

Line 87:

Line 86:

The goal being explored (user-owned, consent-based, portable data) could be approached through several paths, kept here as research references rather than adopted building blocks:

* '''[https://solidproject.org Solid ~~Pods~~]''' (Tim Berners-Lee's model of personal data stores, where each user owns their data and grants applications scoped, revocable access).

* '''[https://solidproject.org Solid]''' ([https://en.wikipedia.org/wiki/Solid_(web_decentralization_project) Wikipedia]), Tim Berners-Lee's model of personal data stores ("Pods"), where each user owns their data and grants applications scoped, revocable access.

* '''[https://holochain.org Holochain]''' (an agent-centric ~~distributed~~ framework where each participant keeps their own validated data ~~and shares it peer-to-peer~~, with no global ledger).

* '''[https://holochain.org Holochain]''' ([https://en.wikipedia.org/wiki/Holochain Wikipedia]), an '''agent-centric''' [https://en.wikipedia.org/wiki/Peer-to-peer peer-to-peer] framework where each participant keeps their own validated data, with no global ledger. Note: Holochain is '''not [https://en.wikipedia.org/wiki/Blockchain blockchain]-based'''; instead of one shared ledger maintained by global consensus, each agent holds its own hash chain and data is shared over a [https://en.wikipedia.org/wiki/Distributed_hash_table distributed hash table].

* '''[https://ipfs.io IPFS]'''~~-style~~ content addressing, and, where genuinely useful, distributed ledgers.

* '''[https://ipfs.io IPFS]''' ([https://en.wikipedia.org/wiki/InterPlanetary_File_System Wikipedia]), content addressing, and, where genuinely useful, [https://en.wikipedia.org/wiki/Distributed_ledger distributed ledgers].

* '''Third-party / federated hosting''', where structured data could live on a trusted third-party server chosen by the user.

Line 101:

Line 100:

=== Selective disclosure ===

For marketplace Transactions, WikiDeal '''is exploring''' a selective disclosure model: parties would share only the data necessary to complete a Transaction~~; reputation scores would be computed locally and verified without revealing the underlying ratings;~~ and dispute resolution could proceed without exposing private communications to the operators unless legally required.

For marketplace Transactions, WikiDeal '''is exploring''' a [https://en.wikipedia.org/wiki/Selective_disclosure selective disclosure] model: the intention is that parties would share only the data necessary to complete a Transaction, and that dispute resolution could proceed without exposing private communications to the operators unless legally required. How reputation could be presented while keeping the underlying ratings private is one of the open questions to be studied; no specific protocol is settled at this stage.

Funders' individual contribution data '''is intended to''' stay strictly confidential. Aggregate figures (total number of funders, total CHF raised, distribution of Community vs [[Gov/en/Portal:Economy/Cash-Rewards|Cash Rewards]] allocations) would be published transparently; individual allocations would never be disclosed without explicit consent.

=== Legal compliance: GDPR, eIDAS 2.0 and beyond ===

This research direction '''is intended to''' be developed in line with applicable data-protection and digital-identity law, not against it.

* '''[https://en.wikipedia.org/wiki/General_Data_Protection_Regulation GDPR]''' (EU General Data Protection Regulation). The principles described above (data minimisation, private-by-default, user consent, purpose limitation, the right to access and erase one's data) are meant to align with GDPR rather than merely tolerate it. Anonymized statistics '''are intended to''' rely on genuinely de-identified data so that they fall outside the scope of personal data.

* '''[https://en.wikipedia.org/wiki/EIDAS eIDAS 2.0]''' (the EU regulation framing digital identity, including the European Digital Identity Wallet). The decentralised-identity work '''aims at''' staying compatible with this emerging framework, so that verifiable credentials and self-presented qualifications could interoperate with officially recognised digital identity where relevant.

These are stated as '''design intentions and starting hypotheses''', to be confirmed with legal counsel as the project matures. Compatibility with other national and sector-specific legislation '''is intended to''' be assessed case by case.

=== Data governance: future hypotheses ===

The following is an '''initial hypothesis''' ~~about digital socio-politics, offered as a projection of possible results in the event of large-scale deployment~~, not as a roadmap.

The following is an '''initial hypothesis''', not a roadmap.

If WikiDeal were deployed widely, data governance might generate broad tendencies within the [[Gov/en/Portal:R&D/Innovations:Rings Of Trust|Rings of Trust]]. It is possible that these tendencies would become a new form of political positioning, much as we have left and right, or conservatism and progressivism. In matters of data workflow, society might one day see, for instance, "libertarians" (favouring maximum user freedom) and "protectors" (favouring stronger built-in safeguards).

Within such a spectrum, it is possible (not certain, and not "will be") that hosting and advertising choices would also become selectable, for example free hosting with advertising, paid hosting without advertising, and community-defined options in between. The guiding ~~hypothesis~~ is ~~that~~ protection ~~would be~~ complete by default, and ~~that~~ users could then choose to ~~reduce some protections~~ in exchange for advantages ~~offered by commercial groups~~ under citizen supervision. The point ~~would be~~ choice, with no single data ~~or advertising~~ regime imposed on everyone.

The guiding idea is simple: '''protection is complete by default''', and users could then choose, transaction by transaction, to share more in exchange for specific advantages, always under citizen supervision. The point is user choice, with no single data regime imposed on everyone. How this choice would be organised in practice is left open, to be defined later in a participative way as the project matures.

----