Overview of the Program

Keynotes

Rigo Wenning, W3C - "DRM and the Web"

Rigo Wenning is working for the W3C as a lawyer in the position of the privacy activity lead & policy analyst. He joined W3C in October 1999 with a focus on privacy and digital signatures. He is also in charge of contact to the European Commission and to European development initiatives.

Andreas Deppe, T-Systems Nova GmbH, "The OPERA Project: Interoperability of Digital Rights Management"

Andreas Deppe works for T-Systems Nova GmbH and is involved with DRM solutions and the OPERA project.

The main subject of Andreas Deppe's keynote talk was the OPERA project that addresses interoperability between various DRM systems and their components. The OPERA project has the objective to specify an open DRM architecture that addresses the needs of all three groups: content providers, DRM system operators, and customers. With respect to rights languages, the OPERA project is using a proprietary rights language. Deppe arguments this design decision with the uncertainty of the REL standards movement, i.e., for OPERA it was not clear which of the current REL standards would prevail. The OPERA project aims at supporting interoperability by issuing licenses that are independent of the DRM system and that are independent of the rendering device. The OPERA project started with the support of a small set of simple business models. More sophisticated ones are planned for a later phase of the project. Deppe stated that DRM systems are not just of interest for the business to consumer market, but also for the business to business market, i.e. for the entire value chain of electronic goods. Finally, Andreas Deppe shared the experience that content providers and service providers currently show a high interest in investing and running interoperable DRM systems.

Willms Buhse, CoreMedia - "OMA Secure Content Delivery for the Mobile World"

Willms Buhse is head of products and marketing and executive board member of the CoreMedia AG in Hamburg Germany. He is also Vice Chair of the OMA Download and DRM working group.

The workshop chairs have invited a representative of the Open Mobile Alliance (OMA) for a keynote talk, as the OMA is an important standards body for the mobile domain and has accepted ODRL as standard rights expression language in its framework. Willms Buhse started his talk with a short introduction of the OMA that has about 370 members and a DRM working group of approximately 50 participants. Willms Buhse states that currently more mobile devices support OMA DRM than the DRM solution of Microsoft. The large acceptance of OMA means at the same time a large acceptance and implementation of ODRL in mobile devices. The OMA standard, version 1.0 which is implemented in the current new phones allow for a "light" DRM, i.e. the implementation of limited DRM functions for low value content. In April 2004 the OMA has completed the OMA standard version 2.0 which provides rights management for high value premium content and includes more sophisticated DRM features as well as improved provisions for security. Buhse stressed that superdistribution of content will play an important role for the mobile domain. Therefore, open standards, i.e. the implementation of OMA standards, are required to enable the distribution between all kinds of different device and service types. The OMA enables the "fair use" of mobile content by supporting user defined domains in which content can be exchanged across devices. He stressed that implementing "fair use" features and interoperability between various DRM devices are vital.

Papers

In this report we only briefly address the main statements of the papers and their relevance for ODRL and its further development. For complete papers (and presentations), please refer to the ODRL Workshop webpage.

Ivana Dusparic presented the paper "A pervasive application rights management architecture based on ODRL". This work is focused on occasionally connected mobile devices exchanging electronic goods and services. As the OMA standard is leading in the mobile domain, ODRL was used for the prototype implementation. Actually, the authors chose a subset of ODRL, using simply the right "execute". The standard vocabulary of ODRL was not sufficient as it does not define rights for functions of software. Therefore, the permission data dictionary was extended. Ivana stated that there was a need to express duties for all contracting parties and that this feature is currently not provided by ODRL. (Please refer to the talk of Susanne Guth for a discussion of duties).

Jaime Delgado presented the paper "Interoperability between ODRL and MPEG-21 REL". He showed an approach how to directly map XrML, respectively MPEG-21 REL to ODRL with the help of XSLT. The audience discussed this approach controversially, as it was lacking a generic information model that mediates between the two languages. Rigo Wenning stated that an approach for the translation from one language to the other requires a mapping of both rights languages to a generic RDF model and then develop an Ontology for it.

Oyvind Vestavik presented the paper "REAP: A System for Rights Management in Digital Libraries". To meet all requirements of the prototype implementation - using the Alexandria Digital Library project - he used an ODRL subset and extended the context element with a specialized library identification field.

Susanne Guth presented the paper "A proposal for the Evolution of the ODRL Information Model". She stated that the current data model of ODRL has some shortcomings, for example, currently rights and duties can only be expressed for the customer but not for the rightsholder, no constraints can be specified for requirements (such as payment within two weeks), etc. Based on the listed shortcomings a new data model was proposed that would enrich the expressiveness of ODRL and eliminate the shortcomings of ODRL. The change of the data model is not an extension of ODRL but a realignment for clear semantics. The paper includes revised XML schemas for the ODRL expression language as well as for the ODRL data dictionary.

Stephen Downes presented the paper "Distributed Digital Rights Management: The EduSource Approach to DRM". Apart from presenting the concept of a "purchaser broker". He was motivating the integration of ODRL into LOM descriptions for learning objects. The rights element of the LOM metadata would link to an extra ODRL document that expresses a certain set of rights. Thus, ODRL rights expressions could be reused by a large number of learning resources. Finally, Stephen argued that there is no "one" DRM system architecture, but different DRM systems for different markets.

Stefan Katzenbeisser presented the paper "Towards Formal Semantics for ODRL". Stefan demonstrated how the formal semantics of ODRL rights expressions can be expressed as automata. Each ODRL rights expression can have different formal semantics. Thus, future ODRL versions should include template automata for each rights expression element that can be combined. Difficulties arise with the automata notation when expressing lend and time-based constraints. Eckhart Koeppen from Nokia commented that for software engineers such formal semantics are very important as formal semantics can be translated to state machines and immensely reduce the risk of misinterpreting licenses.

Olli Pitkanen presented the paper "Nonius: Implementing a DRM Extension to an XML Browser". His team implemented a subset of ODRL in an XML-based browser. They found that ODRL was well written and easy to understand, but implementing some of the enforcement actions, XML lining, and boolean operations was challenging.

Invited Talks

Stephane van Hardeveld presented the talk "Flexible DRM: Real-Life ODRL Implementations". The company VirtuosoMedia in the Netherlands is running several commercial DRM platforms for audio and video content. The company recognized the need for open standards and interoperability. Their DRM solutions are based on ODRL. For the implementations, a subset of ODRL has been used. The customer can choose between some predefined, enforceable contracts formulated in ODRL.

Mariemma Yague presented "An Interoperable and Flexible Infrastructure for DRM". She presented the approach of "semantic access control", i.e. access control that is based on the semantics of the resource and its context. She stressed the need for a general framework capable of supporting very heterogeneous DRM applications and scenarios.

Future Directions

The final afternoon of the workshop was dedicated to looking at the path forward for the ODRL Initiative. The workshop delegates discussed four key areas for future development.

1 - WHY

This area focussed on the high level principles of the ODRL Language that can be used as a fundamental guide for the overall direction and it's objectives. Some of the major outcomes included:

• An open, free, extendable, and simple language.
• Support a broad adoption of the ODRL Language across all sectors.
• Support an information model for contracts which includes rights and duties for all parties in the value chain. This includes offers and agreements/licenses (ie tickets).
• Support the application of rights expressions for free and commercial content.
• Support the use of application profiles of the ODRL language by different sectors including the use of predefined/template contracts.
• Promote the use of ODRL contract expressions as a fair mechanism for binding terms and conditions between parties.

2 - SCOPE

This area focussed on the scope of future versions of the ODRL Language. Some of the major outcomes included:

• Support the use of rights vocabularies from other sectors and communities, such as permissions from the MPEG-21 Part 6 standard and constraints from the ebXML specifications.
• Support the reuse of other metadata vocabularies to supplement the context element.
• Support an integrated approach to identity management as defined by leading industry standards, such as the Liberty specification.
• Express service level agreements within contracts.
• Support the use of templates/models of rights expressions.
• Support flexible negotiations with mechanisms to select/choose different options in offers and agreements and the semantics of such selections.
• Support the application of next rights to an implementable number of levels.
• Clarify the transfer of rights for content that is aggregated/dis-aggregated.
• Support the use of and integration with external privacy standards, such as P3P.
• Provide mechanisms to express common fair use requirements.
• Provide advice on the technical environments required for developing ODRL systems.
• Ensure mechanisms (or transformations) to support backward and forward compatibility of future versions of the ODRL language.

In addition to what may be in scope for ODRL, a number of features of the current version may be removed as they are provided by other standards/systems or have been found to be used infrequently and may overload the language's implementability. Some of these may include, the revoke function, digital signatures and encryption, and entity sequencing.

3 - WHAT

This area focussed on some of the new features of future versions of the ODRL Language. Some of the major outcomes included:

• An updated information model to support contracts in which all parties can have rights and duties.
• Define the semantics using formal notation.
• Mechanisms for wildcards.
• Handling of status information for state-based rights constraints.
• Use of a formal schema registry of ODRL schemas and profiles.
• Mapping to appropriate Ontologies.

In addition, a number of separate documents were envisaged for version 2.0 of ODRL:

• Vision
• Information Model
• Formal Semantics
• XML Schema Binding
• RDF/XML Binding
• Web Services implementation Guide
• Best Practices Guide
• FAQ

4 - PROCESS

This area focussed on the processes and governance of the ODRL Initiative. Some of the major outcomes included:

• Move towards more formal processes to support greater consenus, transparency, and confidence of the ODRL Initiative.
• Investigate potential move to a formal standards body/group and adopt their procedures and processes. Potential groups include; W3C, OASIS, ECMA, IEEE, IETF, CEN/ISS, Open Group. The advantages and disadvantages of such a move will be documented and discussed within the ODRL community.
• Potential to submit current ODRL version to National bodies for adoption.
• Request formal liaison relationship with the OMA.
• Propose and setup an ODRL International Advisory Board to enable governance decisions in a timely and transparent manner.
• Investigate funding opportunities from relevant bodies (eg EU, JISC) to support formal activities.
• Establish Working Groups to undertake the new work required to develop verson 2.0 of the ODRL language.
• Utilise ODRL website, email-lists, archives, and document repository to manage the working groups activities.
• Review the ODRL Initiative Hosts, Supporters, and membership program.
• Promote external ODRL activities and support communities developing ODRL profiles.

Next Steps

The Workshop was clearly a major success and provided the timely and necessary forum for ODRL researchers and implementors to share experiences and new ideas.

The key next steps are to develop a work plan for the technical development of the next version of ODRL and to establish the ODRL International Advisory Board to manage the goverance processes for the Initiative.