Machine Readable Rights Document

Rights Expression Working Group
https://www.flickr.com/photos/75487768@N04/12188256115
Stuart Myles * Associated Press * 22nd October 2014
© 2014 IPTC (www.iptc.org) All rights reserved

The News Industry Needs
Machine Readable Rights
Increasing automation of workflows
Less opportunity for editors to read notes,
the traditional means of indicating restrictions and duties
Sophisticated publishing relationships
Fewer single purpose, static connections
More (need for) flexible, ad hoc uses of content which
respect rights
© 2014 IPTC (www.iptc.org) All rights reserved 2

RightsML
Founded on ODRL
Open Digital Rights Language
Selected by IPTC after evaluating several alternatives
http://www.w3.org/community/odrl/
RightsML is a news industry specific vocabulary
Plugs into ODRL framework
http://rightsml.org
http://dev.iptc.org/RightsML
Machine processable permissions, restrictions and duties
Derived from media industry requirements
Can be embedded within content or stand alone policies
Designed for all media types
Designed for automatic evaluation

Illustrating the Need for RightsML
• Let’s say that I am a publisher of an Internet magazine
• My magazine features a single cat photo each day
– Cats are popular on the Internet
• My publishing company is based in Paris, France
• My supplier of cat photos is the Example Photo Agency
• The EPA licenses me to publish their photos on my site

THE DAILY CAT
cats.example.com THE WORLD’S FAVOURITE CATS - Since 2014
I GET A CAT PHOTO FROM EPA
EPA sends me a cat photo
My license with the EPA
says I can publish their
photos to my website.
There is no additional policy
associated with the photo.
So, I publish the photo on
my website.
Le chatte by FLASHFLOOD https://www.flickr.com/photos/flashflood/7299891846

USE WHEN
LOCATION ISN’T FRANCE
The next photo I receive from EPA has a policy
The EPA sends me a photo with a RightsML policy:
Policy
Use when location is not France
So, even though my EPA license
says I may publish their photos
to the web
The RightsML policy for this one
photo overrides it
(As I am located in France)
Mona Lisa lolcat by Michael Russel https://www.flickr.com/photos/planetrussell/6814444991

THE DAILY CAT
cats.example.com THE WORLD’S FAVOURITE CATS - Since 2014
PHOTO: “NO USE UK” POLICY
Policy
Use when location is not UK
NO USE UK. There is a
policy attached to this photo:
“Use when location is not
UK”.
Since my license from the
EPA says I can publish their
photos to my website and
my website isn’t based in the
UK, I can publish the photo
on my website.
Note, however, that the
policy attached to this photo
must be interpreted in the
context of the license from
the EPA. Just because the
policy says “Use when
location is not UK”, it doesn’t
mean that other uses are
granted by the policy. So, for
example, I can’t create print
the photo, as my EPA
license only grants web use.
tea & cats by Kayla Nicole https://www.flickr.com/photos/ohpapercut/14801702282

Permissions, Restrictions, Duties
• RightsML covers more than just geo restrictions, e.g.
– Geography
– Time
– Distribution channel
– Default “read the editors’ notes”
– Action
– Duty to pay
– http://dev.iptc.org/RightsML-Examples-Simple-Cases
• And more than just one simple constraint
– Using the “simple” examples as building blocks
– http://dev.iptc.org/RightsML-10-Implementation-Examples
• Many licenses and policies are much more complex

Traditional Publishing
A feed per publication
A tuned
content
set
Licensed
for one
outlet
Editors
review
notes
Often by
media
type

One Publishing House
With many publications
Content still
licensed per
outlet
Newspapers,
magazines,
broadcast
channels
Websites
and apps for
desktop,
smartphone,
tablet
Content
duplication
Enshrines
legacy
relationships

Most Publishing House
Take Content from Multiple Providers
Lots of
complexity
and waste for
publishers and
providers.
Harder for
publishers to
respond to
new
opportunities

RightsML Enables Automated
Publishing That Respects Rights
Content still
licensed per
outlet
Automatically
route content
Less editorial
intervention to
check
No content restrictions
duplication
Apply
restrictions
per content
item
New uses for
content without
a duplicate
feed

RightsML Processing Model

Contract Processing Model

Policy Processing: Permissions

Policy Processing: Restrictions

Why the recursion?
• ODRL (and hence RightsML) has a hierarchy of actions
– http://www.w3.org/community/odrl/work/2-0-common-vocabulary-
constraint-draft-changes/#section-22
• For example, a desired use is "print", which has a parent
of "present", which has a parent of "use“
• A given license might permit "present" and thus also
permit the child actions of "display", "play" and "print"

Software Supporting ODRL v2
• licensed
– https://github.com/iptc/rightsml-dev/tree/master/licensed
– Python library to read and write ODRL 2.1 and RightsML
– Supports XML and JSON encodings
• rightsml-lib-js
– https://github.com/iptc/rightsml-dev/tree/master/rightsml-lib-js
– Javascript library to create ODRL 2.1 and RightsML
– Supports XML and JSON encodings
• odrlapi
– http://vroddon.github.io/odrlapi/
– Java library to read and write ODRL 2.0
– Supports RDF encoding

Refining ODRL
• Working with ODRL W3C CG
– Simplify, enhance ODRL
– Improve documentation
• ODRL 2.1
– Added profile attribute, datatype and unit to constraints
• http://www.w3.org/community/odrl/work/2-0-core-model-constraint-draft-
changes/
– Slimmed down the Common Vocabulary
• http://www.w3.org/community/odrl/work/2-0-xml-encoding-constraint-
draft-changes/
– JSON http://www.w3.org/community/odrl/work/json/
– RDF http://www.w3.org/ns/odrl/2/
– XML http://www.w3.org/community/odrl/work/2-0-xml-encoding-constraint-
draft-changes/

Next Steps
• Finalize ODRL 2.1
– And publish updates to RightsML documntation and specs
• Proof of concept implementation of the processing model
– Demo POC based on licensed in NYC
• Outreach
• http://dev.iptc.org/RightsML
• iptc-rightsml-dev@yahoogroups.com

Date and Place of Next Meeting
New York, USA March, 2015
https://www.flickr.com/photos/arnebornheim/8080932433
Danke und auf Wiedersehen!

Machine Readable Rights Document

Recommended

Recommended

More Related Content

Similar to Machine Readable Rights Document

Similar to Machine Readable Rights Document (20)

More from Stuart Myles

More from Stuart Myles (20)

Recently uploaded

Recently uploaded (20)

Machine Readable Rights Document