[p2p-research] About Right Expression Languages

[Hervé Le Crosnier]

	Hello,

	After is a good introduction to Right Expression
	Languages, i've taken from a canadian journal.

	What i've told about ODRL is not to blame all
	REL, even then ODRL... I think we need some kind
	of things for any user (even a robot) to know	
	what he/she(it) can do with a document. The good example
	is the RDF version of the Creative Commons licence.

	We also need a complete metadata set to really know
	who is playing (singing, videoing, writing,...) into
	any piece of digital document. Certainly through
	a unique identifier with associated metadatas in
	some decentralized mesh of servers (like the DNS).

	But interpretation of those informations can't rest on
	computer program. You can say this digital document
	is following such and such format, or such and such
	guideline... but saying "you need that software to
	use such a document" or "we control what you do with the
	document and automatically stop any use complaining
	with our policy" (which is the very core of DRM systems)
	is another step for destroying what a society (as a
	complex system) is.

	Why do we respect law we never read ? Here is some core
	aspect of a society. Saint-Just, a french revolutionnary
	of the 1789, says to deputies : "give us law we will love".

Hervé Le Crosnier


http://www.osbr.ca/archive.php#A4
An Introduction to Rights Expression Languages


"Language is a process of free creation; its laws and principles are
fixed, but the manner in which the principles of generation are used is
free and infinitely varied." Noam Chomsky

The objective of this article is to: (i) extend the discussion of
licensing to non-software assets and (ii) provide an introduction to
rights expression languages (RELs). Licensing is not limited to
software. We can associate a license with any kind of asset that holds
intellectual value, and can thus be turned into a source of revenue.
Here, our interest is on information assets, which include software and
software components, but also services, processes, and content. For
instance, a song that a user downloads from iTunes is an information
asset. So is a web service such as the Google Maps API (application
programming interface).

Licensing and DRM

We begin the discussion by defining two key terms, licensing and digital
rights management (DRM). Licensing is a fundamental way of controlling
the distribution of information assets, and underlies the design of
business relationships and strategies. Licensing principles reflect the
overall business value of assets to their producers and consumers. Also,
licensing is often used to protect the intellectual property rights
(IPR) of the producers of an asset. Licensing is both a source of
revenue and a strategic tool.

In their book Digital Rights Management: Business and Technology,
Rosenblatt, Trippe, and Mooney define DRM as an umbrella term referring
to the collection of technologies (hardware, software, and services)
that govern the access to information assets through associated rights,
and controls their distribution. The foundation of DRM technology relies
on our ability to represent the rights over digital assets. RELs
represent the rights over assets in a machine-understandable way. RELs
describe different aspects of usage control, payment, and access, for a
digital access environment.

According to Parrott, a REL consists of four components:

    * Subjects, the actors who perform some operation or action
    * Objects, the content against which a subject wants to perform an
operation
    * Operations or what the subjects wants to do to the object
    * A set of constraints or conditions under which an operation can be
performed

These components and their relations support a range of models, each
describing a way of applying digital rights. In general, a REL expresses
the rights of an information asset either in some form of logic or in an
XML-based language.

To illustrate these concepts, let us assume that a user wants to
download a song from the iTunes store and play it on his iPod. The
subject is the user, the object the song, the operation to play the
song, and the constraints are that the user has to pay 99 cents for the
download and cannot share the song with his friends.

Rights Expression Languages

What follows is a brief history of RELs. A pioneering formal language
called DigitalRights describes a mathematical model of simple licenses
that consists of payment and rendering events and a formal
representation of licenses. LicenseScript is a logic-based REL.

Logic-based RELs express general prepositions of a permissive or
obligatory (restrictive) statement. However, these languages cannot
express a finer level of granularity of the assets, actors, or actions
involved. Logic-based RELs cannot interoperate with other types of RELs.

XML-based RELs support interoperable ways of expressing the rights of an
information asset. An XML-based REL allows asset producers to specify
flexible expressions. The Extended Rights Markup Language (XrML) and the
Open Digital Rights Language (ODRL) are two XML-based RELs which have
gained international recognition and are widely used in industry.

XrML is the basis for the REL of the MPEG-21 multimedia framework. It
focuses on the license through which a rights holder confers usage
rights to a consumer. A license can be digitally signed by the rights
holder, now also referred to as the issuer, to confirm that the holder
grants the rights contained in the license. An XrML license contains one
or multiple grants and the license issuer. A grant is the element within
the license that authorizes a subject to exercise a right on some object
under some constraints. Note that the actual terminology used by XrML is
slightly different from this. ODRL is an open standard language for the
expression of terms and conditions over assets in open and trusted
environments. ODRL consists of an expression language and a data
dictionary. The expression language defines basic terms of rights
expressions and their organization using a set of abstract concepts. The
data dictionary defines the semantics of the concrete terms used to
express an instance of a rights specification.

ODRL is based upon an extensible model for rights expression, and
defines the following three core entities and their relationships:

    * Assets, the objects being licensed
    * Rights, the rules concerning permitted activities, the constraints
or limits to these permissions, the requirements or obligations needed
to exercise the permission, and the conditions or specifications of
exceptions that, if true, terminate the permissions and may require
re-negotiation of the rights
    * Parties, the information regarding the service provider, consumer,
or broker

With these entities, ODRL can express offers (proposals from rights
holders for specific rights over their assets) and agreements (contracts
or deals between the parties with specific offers). ODRL supports the
declaration of a wide range of expressions. It can also be extended to
different types of domains. For example, we can use ODRL to specify that
a consumer of a geocoding web service can only use this service in a
non-commercial context, as well as the number of times the service can
be accessed each day. ODRL has been published by the World Wide Web
Consortium (W3C), and has received wide acceptance. ODRL is supported by
several industry consortia such as the Dublin Core Metadata Initiative
(DCMI) and the Open Mobile Alliance (OMA).

Two applications of ODRL are an ODRL profile of the semantics of
Creative Commons (CC) licenses and the ODRL profile for services
(ODRL-S). The core semantics of CC licenses have been expressed in ODRL.
This profile supports extensions to these semantics, and defines an XML
Schema. ODRL-S is an extended version of ODRL to express clauses for
service licensing, creating a machine-understandable service license.

Conclusion

Information assets are usually accompanied by a license that describes
the terms and conditions on the use of this asset imposed by its
producer. A license reflects the overall business value of the asset to
its producers and consumers. The kind of rights vary based on the nature
and context of the assets involved. For example, one of the rights for a
multimedia asset is that consumers can play it. The concept of playing
can not be directly applied to a web service asset.

Similarly, the rights governing the use of the interface and
implementation of a web service are distinct. However, for multimedia or
software assets we cannot make such a distinction. In this article we
introduced the concept of licensing and RELs, and briefly described XrML
and ODRL as the two most prominent RELs.

We thank Dr. Renato Iannella, NICTA, Australia, and Prof. Vincenzo
D'Andrea, University of Trento, Italy, for their suggestions and comments.

G.R. Gangadharan is a doctorate student in University of Trento, Trento,
Italy. His research interests include Free/Open Source Software Systems,
Service Oriented Computing, Internet Software Engineering and Web 2.0,
and Business Models of Software and Services.

Michael Weiss holds a faculty appointment in the Department of Systems
and Computer Engineering at Carleton University, Ottawa, Canada, and is
a member of the Technology Innovation Management program. His research
interests include open source ecosystems, service-oriented architectures
and Web 2.0, business process modeling, social network analysis, and
product architecture and design. Michael has published on the evolution
of open source communities and licensing of open services.