Structured Commons

Definition

A certificate of existence (CoE) binds an object with a date stamp and asserts that the object existed no later than the certified date.

A CoE exists outside of the object that it certifies. Each object can be attested by more than one CoE.

CoEs are intended for display in citation contexts, for example "this document refers to another document D, attested by CoEs X, Y, and Z".

The Structured Commons model supports multiple certification methods to generate and verify CoEs, including a variety of historical mechanisms already in use in academic publishing.

Certification methods

New certifications methods may be defined over time, as long as they are publicly documented and certificates can be independently verified by users.

Each certification method must provide at least the following services:

• from an object or fingerprint, generate a CoE for a calendar date no earlier than the date of the request;
• from a CoE and object, verify the CoE is valid and return the earliest attested date for the fingerprint;
• represent the CoE as a textual, printable CoE signature that can be embedded in citation contexts.

All CoE signatures must match the following common format: the method name, followed by a colon, followed optionally by the certified date stamp with format YYYY-MM-DD and a colon, followed by a textual representation of the CoE:

signature: METHOD ':' [ DATESTAMP ':' ] CERTIFICATE

This format is purposely defined to overlap with the syntax already used in academic works to refer to certificates of existence, cf. the next section.

Predefined certification methods

The following CoE methods are predefined:

These methods correspond to historical mechanisms that are already trusted by scholars to deliver reliable registration services.

Some example valid CoE signatures:

arxiv:2014-04-30:1404.7753

method: arXiv, attested date April 30th, 2014, certificate "1404.7753"

arxiv:1404.7753

method: arXiv, omitted date, certificate "1404.7753"

doi:2014-06-09:10.1145/2618137.2618139

method: DOI, attested date June 9th, 2014, certificate "10.1145/2618137.2618139"

doi:10.1145/2618137.2618139

method: DOI, omitted date, certificate "10.1145/2618137.2618139"

acmid:2014-06-09:2618137.2618139

method: ACMID, attested date June 9th, 2014, certificate "2618137.2618139"

acmid:2618137.2618139

method: ACMID, omitted date, certificate "2618137.2618139"

Hypertext references

In a hypertext context, each CoE signature representation should be configured as a hyperlink to the certificate URI.

For example, an arxiv CoE sig can redirect to the URI http://arxiv.org/abs/ followed by the arXiv identifier: arxiv:2014-04-30:1404.7753

See the table above for a list of base URLs to use for the predefined certification methods.

Very long term durability

As certification requires trusted 3rd parties, a risk exists that the disappearance of a certification authority will make the verification of a single internet-based CoE difficult, expensive or even impossible.

This risk increases over time, and ensuring CoE verification stays possible is quite relevant considering the Structured Commons model envisions preserving knowledge over the very long term (50+ years).

Historically, durable CoE verifiability was ensured by printing journals on paper, and archiving multiple paper copies over time and in different libraries over the world. Even hundred years later, we are able to check 18th century articles existed no later than a particular date by investigating and cross-checking paper archives. This method is effective even for works that are not published in journals, as long as journals cite them: if a work A is cited by journal J which is attested to exist at date D by a set of printed copies, then A is attested to have existed no later than D.

Starting with the Internet, more and more works are published "online only" and never cited by printed journals, so the CoE durability provided by journals becomes largely ineffective.

There are three strategies to guarantee the very long term verifiability of CoEs, even without printed journals, in increasing order of effectiveness:

1. creating multiple CoEs using more than one certification method (decreases the risk that the disappearance of one certification authority makes verification impossible);
2. archive CoE databases in durable media (eg. paper or microfilm archives that map fingerprints to known CoEs);
3. ensure that works are cited using both their fingerprint and known CoEs, so that the structured global web of knowledge serves as distributed trusted database of CoEs.

The last strategy is the most robust: if a work A is attested by CoE X, and B cites A using A’s fingerprint and X’s signature, and the web of knowledge attests B via fingerprint by citation in multiple other works, then it becomes impossible to modify A or fake another CoE (or attested date) because this would change A’s fingerprint and thus B’s. Over time, the structured citation network will become the trusted database of CoEs for past works.

References