A controversial position, but this note explores what is needed in a digital rights expression language (DREL) for learning technology systems and questions what a DREL adds to existing learning technology data models and languages. Many of the ideas and concepts behind the rights, roles, constraints and workflow or sequencing representations present in current DRELs (e.g., ODRL, XrML) are already present in learning technology systems and associated standards. What else do we need for digital rights management in learning technology, and how should we express it?

DREL Concepts

DRELs and DRM systems concepts include:

• Rights: permitted actions, e.g., print a learning object, view a lecture.
• Constraints: limit conditions on use, e.g., use learning object on certain days, in certain places, so many times.
• Requirements: required actions and pre/post conditions on actions, e.g., register (and pay) for course before use, track use.
• Conditions: combinations of constraints and permissions, i.e., you can do "x" subject to constraints and requirements "y".
• Parties: descriptions of entities involved in DRM, e.g., learner, author.
• Offers and Agreements/Licenses: aggregations of conditions and permissions between parties.
• Security: digitally signed content (relying on existing models and representations).

DREL/DRM Use

DRELs and DRM systems are used to express and enforce:

• Controls and conditions on the use of a content object in the form: user "x" may do "y" with content under condition "z". Conditions may be relatively complex, requiring general logical combination of individual constraints and overall sequences of actions.
• Trading descriptions, i.e., agreements to a particular set of conditions between two parties.

DRM examples are often described in terms of DRM enforcement being part of a closed system, e.g., a media player limits access to the content within the conditions expressed by the DREL, a content repository enforces constraints when requesting content.

DRELs simply express rights and conditions and expected behaviors on any conforming use by any supporting technology system. Behavioral controls (e.g., enforcement) is separate from rights expression.

Learning Content Management Concepts

Many of the concepts present in digital rights expression and management are similar to those used in learning content management, particularly learning delivery.

• Activities: user interactions with associated learning content resources, e.g., render the content, take assessment, grade assessment, participate in a discussion, annotate a work.
• Limits: limits on content use, e.g., when, where, how many times, device.
• Sequences: ordered steps, e.g., pretest before learning, remediation following assessment, access following authentication and authorization.
• Conditional Use: combinations of limits, sequences and activities.
• Roles: individuals or groups with associated rights.

Learning technology systems combine these concepts to control the learner experience: learner (user) in role "x" may do activity "y" with content object under use, limit or sequence condition "z".

In the learning context, using learning content is more than just controlling access to the content.

Similarities in Digital Rights and Learning Content Managements

How are the concepts of DRM and a DREL special? Learning technology systems apply roles, rights and constraints to when learning can take place and how learning activities are ordered and presented. The core concepts for controlling learning are similar to those of content rights management.

• Activities & Rights: same concept, different overlapping vocabulary of terms.
• Constraints & Limits: same concept, different overlapping vocabulary of terms.
• Requirements: same concept, different overlapping vocabulary of terms.
• Conditions: same concept, different overlapping vocabulary of terms.
• Parties & Roles: same concept, different overlapping vocabulary of terms.
• Offers & Licenses: no direct overlap, unique to DRM.
• Security: same need, not yet part of formal content and learning standards (assumed to be part of a lower-level infrastructure).

Learning technology standards and data models are used to describe expected behaviors of conforming systems and separate the expression from the implementation used to achieve conformance.

Standards and DREL

Learning technology standards contain elements of all the overlapping areas, e.g., Sequencing, EMLs and QTI express activities, conditions, rules, constraints; LIP describes groups and roles. Workflow (the generalization of conditions, sequences and limits) and security exist in both learning-specific standards, e.g., the OKI APIs, and in more general-purpose standards, e.g., WSFL, SAML.

DREL and the learning technology standards listed also share a common feature: they express behaviors, i.e., how systems are expected to process learning content and rights expressions. These standards are not just data formats for moving static data sets between systems.

So What's the Problem?

There are similar and overlapping behaviors expressed in both DRELs and different learning technology standards. Should we develop or adopt a new DREL standard for learning systems? Should we extend the existing learning technology standards to include the requisite DRE vocabularies and rights behaviors? Is there a middle ground?

Consider the following simple example of a naive implementation of DRM (expressed using a DREL) in a learning content delivery system, e.g., an LMS/LCMS.

• A learning designer has specified a set of limits and sequences on content, e.g., (1) the learner must view a lecture (a content object) and then take an assessment (a different content object), (2) the learner may review the lecture an unlimited number of times, (3) the learner is limited in how many times he may take the assessment. These are expressed in some learning experience language.
• There is a trading agreement with the content author that specifies the number of times the lecture and assessment may be accessed, expressed in some DREL.
• The learner (already authenticated to the learning delivery system) requests one of the content objects.
• The learning delivery system evaluates the learning designer's conditions on use of the content. The conditions are met, and the learning delivery system requests the content resource from the content repository for display through a media player.
• The content management system applies the DREL conditions and determines that the request cannot be honored.

What happens next? Does the learning delivery system capture the failure and react or does the learner get an error directly from the media player? Have we confused the learner by exposing similar types of behavior limits at multiple points in the learning experience and through multiple components?

By splitting the management of the behavior into different system components and representations, we have introduced the complexity of coordinating a consistent behavior and user experience across the components.

In an alternative implementation, the learning delivery system would use the DREL expressions in addition to the learning designer's intention to decide if the learner's request is supported. The DREL specified behaviors could be intermixed (logically) in the processing. The learning delivery system must understand both the learning experience language and the DREL, and be able to control both the learning experience and manage digital rights. In this approach, we might not need a completely new DREL, but we need to incorporate DREL expressions and vocabularies into current approaches.

Alternatively, we can let the content tools manage only the digital rights. This requires a single, standard mechanism for any learning technology system to access any content management system to validate requests against the rights. In this approach, what matters to the learning technology system is how it communicates with the content system, not the details of the DREL used within any content system. Do we just need a DRM service specification so we can ignore whatever DREL is used?

Conclusions

We are beginning to see the problems resulting from developing isolated solutions around specific needs, e.g., rights management, content sequencing, without understanding that there are some common underlying fundamentals, e.g., workflow, constraints, security. When we mix the different solutions and their different approaches and representations, we find integration difficult.

DRELs do have some unique characteristics, i.e., trading descriptions, but much of what they provide for expressing rights control overlaps with general requirements for content management within learning technology systems. DRM should not be an add-on feature, or just delegated to content repositories and media players, but fully integrated into overall learning content management. Unless properly integrated into existing learning technology systems, digital rights management will complicate the development and delivery of learning content.

The challenge is to describe digital rights expressions in a way that integrates with existing models. We should not avoid asking: How do we seamlessly integrate a DREL or DRM solution into learning technology?

Update: 15062002
© Copyright 2002, Carnegie Mellon University
All Rights Reserved