(Edit after first answer to avoid cross-post)

Popular public copyright licenses like GPL3, MIT, Apache2, CC-BY-SA3, etc. have official human-readable texts, and, I suppose, in nowadays, have also translations to "machine-readable texts", that are REL interpretations/translations of the official texts... Each license is a set of clauses that can be described by REL, and this set is my interest.

There are a "license catalog" (content with semantic markup) with each license translated to REL or RDF?

The RDFLicense dataset (with git here and an article here) is a near to perfect catalog (!), but have some critical problems:

  1. No criteria for grouping "semantically equivalent licenses", that is, equivalent versions (where no relevant "contractual change" was made) and equivalent translations (by definition an translation must be equivalent). So there are two problems: 1.1) no criteria; 1.2) the catalog has very less than 100 "real licenses", then its volume is yet small.
    PS: see semantic definition of frbr:Expression, frbr:realization, etc.

  2. No standard markup to show content with semantics as microdata examples, that have "probative value" for indicated/elected properties of the license.

  3. Use of "exotic" standard in both, the license-content (used VoID description) and summarized (only RDF-like, used .ttl) dataset.
    PS: the ideal is pure RDFa (a W3C de jure standard) or Microdata (most popular, a de facto standard) for content-semantic markup; and JSON-LD for summarized semantic (and join efforts with OKFN project or other iniciatives).

So, a good suggestion to overcome these problems is also a good answer.


Existing translations and samples. The only that I found was a CC licenses translated to CC-REL.

"REL is RDF". Any REL (Rights Expression Language), ex. CC-REL or RightsML, can be translated to RDF, so we can suppose that any popular license tranlated to REL, is a RDF descriptor of the license.

"RDF is JSON-LD, RDFa, Microdata", no problem for me, all can be translated to another.

"catalogues". They exist, as Wikipedia lists, OKFN lists, tldrlegal lists, and others... But no one have a REL or RDF translation/interpretation.

About REL lanuages and license catalogues, Rodríguez-Doncel et al. (2014) commented,

(...) languages to digitally represent the key information in licenses have existed for at least a decade (...), but no effort was made to systematically map existing licenses to these languages.


Reference-example (BSD-2-Clause license) of markup text,

Redistribution and use in source and binary forms (...) are permitted provided that (...):

  1. Redistributions of source code must retain the above copyright notice (...).
  2. Redistributions in binary form must reproduce the above copyright notice (...).

So, in HTML + Microdata markup, will be something as

<div itemscope itemtype="http://creativecommons.org/ns#License"> 
<p itemtype="#Permission"> Redistribution and use in source and binary forms (...) are permitted provided that (...)</p>
<p itemtype="#Distribution">1. Redistributions of source code <span itemprop="#notice">must retain the above copyright notice</span> (...).</p>
<p itemtype="#Distribution">2. Redistributions in binary form <span itemprop="#notice">must reproduce the above copyright notice</span> (...).</p>

and summarizing it in JSON-LD something like,

  "@context": {
  "@id": "cc:License",
  "cc:Permissions": [ 
    "@type": "cc:Distribution",
    "value": "Redistributions of source code",
    "cc:notice": "must retain the above copyright notice"
    "@type": "cc:Distribution",
    "value": "Redistributions in binary form",
    "cc:notice": "must retain the above copyright notice"

work and information chain

The "marked license" (example with Microdata), is the ideal aim, and need a lot of human work (!), reviewers and endorsement to be valid as oficial interpretation.

The "summarized license" (example with JSON-LD) can be obtained from the markup by software; and a "basic extract" like this (columns "is_by", "is_sa" and "is_nd") can be obtained by software from the "summarized license".

1 Answer 1

  1. The criteria for grouping "semantically equivalent licenses" might be implemented as an external HTTP REST service. There are already some services based on that dataset running here: http://licensius.com/apidoc/index.html

  2. Volume of licenses. Contributions of new licenses or suggested revisions to this github folder are welcome; only a TTL has to be dropped for each new license here. This might be a good tool against the "license proliferation" problem.

  3. Yes, having microdata would be great. Deep linking for each of the asserted statements would be a plus (namely, if the license claims "distribution is permitted" then a link to the specific paragraph would be great.

  • Hi Victor is an honor to talk with the author of the cited best work of this issue (!). Good API! I edited again: see examples and "information chain". About item 1 (equiv. licenses) of your answer, I was thinking in the work (human time) to describe or check each license... Is only a normalization problem, imagine that each "license as contract" have an ID, so translations and other variations are respect to the same ID... Other information is redundant, and can be listed here. Feb 15, 2016 at 2:11
  • About item 2, thanks (!) I will try, and we can discuss something more at issues. About item 3, ok... I am thinking how to convert license-text into marked text, in a collaboration context perhaps store and edit license-texts at Mediawiki engine is a solution. Feb 15, 2016 at 2:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.