Managing electronic resources
The terms "Digital Rights Management" and "Rights Expression Language" are somewhat unhelpful, often interpreted as having more to do with technical protection and limitation of access than the management and expression of rights. This article, however, addresses the very real issues of how digital rights are to be managed and expressed in an environment of trust – the licensing relationship between academic publishers and libraries.

In this business to business relationship technical protection is neither required nor desirable. The relationship is a contractual one, based on licences and publishers are quite content to trust librarians to comply with the terms of those licences. Libraries may decide to incorporate some degree of technical protection within their own library systems in order to prevent unauthorised use of digital materials, but they do not want to have this imposed by the publishers.

However, as the number of digital resources in library collections grows, libraries are reaching out for ways to help them comply with the differing licensing terms applied to resources by their creators and publishers. Paper licences, once negotiated and exchanged, are duly filed and it becomes a major task to answer a simple question such as "May I make 20 copies of this article for my class?" The ability to express usage rights and permissions electronically in a simple form, link to them from digital resources and communicate them to users has become an urgent need. Ideally, an XML message expressing the terms of the publisher/library licence should be generated by the publisher’s licensing department and communicated to the library, either directly or through a trusted intermediary, for linking to the relevant resources.

Neither publisher nor library systems have been able to cope with this requirement. One significant obstacle has been the lack of underlying metadata standards necessary for such complex exchanges.

Work on rights metadata standards
EDItEUR (cf. sources) is the international body for e-commerce standards in the book and serials sector, originally set up by the European Bureau of Library and Documentation Associations, the Federation of European Publishers and the European Booksellers Federation. It is now a truly international not-for-profit organisation with members from all of the above communities in USA, Canada, Europe, Asia and Australia developing standards for EDI, bibliographic information and the communication of serials information.

EDItEUR has been working in the area of rights since 1997, when a joint NISO (National Information Standards Organisation; cf. sources) /EDItEUR working party was established to explore the metadata requirements for rights trading. The working party concluded that the essential elements that had to be identified and described were the resource itself, the user and the required use. Progress has been made in the area of resource identification, with the widespread implementation of DOI (Digital Object Identifier; cf. sources) and open URL to identify resources and the development of ONIX (Online Information Exchange; cf. EDItEUR 2005a), and in particular ONIX for Serials, to describe them. The identification and description of users raises privacy as well as administrative issues and, in actuality, the major requirement is not individual identification of users but authentication of their status as a bona fide user. Authentication systems such as Shibboleth (cf. sources) now enable authentication of user status.

There remains the description of usage rights. So-called Rights Expression Languages such as XrML (eXtensible rights Markup Language; cf. sources) and ODRL (Open Digital Rights Language; cf. sources), have been developed primarily for the music and video industries as mark-up languages to drive technical rights enforcement technologies in business to consumer situations. They are, at the same time, both more and less than what is required for the communication of licensing terms information.

EDItEUR believed that what was required, and lacking in the rights expression languages, was a highly-structured data dictionary that accommodated the full richness of licensing terms and that could be used to generate messages in various different syntaxes. On the other hand, there was no requirement in this community for messages to directly drive technical protection mechanisms.

This work was picked up in a multi-media context by the EU-funded (Interoperability of Data in e-Commerce Systems; cf. sources) project that ran from the end of 1998 to early 2000 and included participation from the various media sectors. developed an analysis of the requirements for metadata for e-commerce in intellectual property in the network environment which received widespread support. The work of the project has been carried forward in the development of the iDD (indecs Data Dictionary) by the International DOI Foundation (cf. sources) and in the development of the ISO MPEG Rights Data Dictionary - ISO/IEC 21000-6 (cf. Barlas 2005).

The Electronic Resource Management Initiative
Parallel to these developments, the library community was struggling with the wider issues of managing electronic resources and given the inability of integrated library systems to provide comprehensive solutions at the time, individual institutions were developing their own non-interoperable systems. Recognising the shortcomings of such an approach, in mid-2002 the Digital Library Federation launched ERMI (cf. sources), the Electronic Resource Management Initiative to define the functional requirements of an electronic resource management system and begin to develop a common set of specifications which could be followed by the library systems vendors.

ERMI’s goals were to:

  • Describe architectures needed to manage large collections of licensed e-resources
  • Establish lists of elements and definitions
  • Write and publish XML Schemas/DTDs (Document Type Definitions)
  • Promote best practices and standards for data interchange

These goals were substantially achieved in their final report (Jewell et al. 2004) published in August 2004. Many of the major library systems suppliers have already started building electronic resource management extensions to their systems based on the ERMI reference model.

A number of very complex issues were recognised and modelled, including the complex ways in which licences relate to electronic resources; the need to interpret the lack of a specific statement in licenses (where does such silence imply a permission and where a prohibition); the difficulty of managing the realities of complex user groups and institutional locations.

ONIX for licensing terms
Meanwhile, EDItEUR, had been working in a Joint Working Party with NISO to develop "ONIX for Serials" a family of XML standards to support communication between publishers, agents and librarians primarily as it relates to the management of e-journals.

A requirement for the unambiguous electronic communication of licence terms within this supply chain was identified and EDItEUR commissioned a paper from the Rightscom consultancy (cf. sources) (An assessment of ERMI in the context of ONIX and requirements for recording and communicating licence terms for electronic resources (Bide and Rust 2004), to assess the extent to which the ERMI work in this area might provide a basis for the development of standards for the transmission of licensing terms throughout the supply chain for digital resources.

The assessment paper concluded that the ERMI work was a good starting point for such work but would require further development in order to meet EDItEUR’s requirements that a licensing terms message should:

  • Take into account the requirements of all stakeholders in the supply chain: libraries, publishers and other rights holders, intermediaries, library users
  • Provide for the full complexity of rights expression:
  • Be designed to support interoperability
  • Be fully extensible in future, to support new business models, all types of use and all media types

The paper recommended the development of a generic ontological structure for rights based on a "contextual", event-based architecture and a well-structured rights data dictionary. This proposal was presented at a seminar on the subject jointly hosted by EDItEUR and NISO in London in December 2004. The feeling of that seminar was that a proof of concept was required to better illustrate the potential of "ONIX for Licensing Terms".

Following the seminar, with funding from the Publishers Licensing Society (PLS; cf. sources) and the JISC (The Joint Information Systems Committee; cf. sources), Rightscom were commissioned to undertake this "proof of concept" project, working with the EDItEUR ONIX technical team (David Martin and Francis Cave) to explore the possibility of developing an "ONIX for Licences" message that could be used by publishers and online hosts to communicate licence terms to libraries and subscription agents (EDItEUR 2005b).

The aim of the project was to produce a prototype XML message for communicating in a computable form the terms of a Licence agreement for the use, by libraries, of a publisher’s digital works. The main use case was the licensing of electronic journals, but the structure of the message needed to be flexible enough to be extensible to any other type of digital media and license in future by adding to its semantics but not significantly changing its structure. The message therefore needed to be generic in structure but successfully demonstrate an initial, specialized application.

The prototype message was produced as an XML schema and succeeded in demonstrating that each element of the example licence clauses could be fully modelled. The modelling also highlighted the range of possible variations within even apparently simple licensing clauses, and the limitations of the original ERMI approach that defines only a limited "typical" set of usages with no mechanism for variation. One example of this is the use of the term ILL (Interlibrary Loan), frequently used in library licences.

ILL is not a single permission – it is a complex bundle of permissions, prohibitions and conditions with many variables. These can be expressed in a very generic form – e.g. permits "ILL"; or in a very granular and complex form, e.g.
permits a librarian at institution "A" to make a copy of a defined part of resource "X" in physical (but not digital) form and sending that copy of part of resource "X" to a librarian at another institution "B" – subject to the condition that institution "B" is in the same country as institution "A" – and then the librarian at institution "B" may pass that copy of part of resource "X" to a user – subject to the condition that the user is an employee of institution "B" and is using the copy for academic non-commercial research – and all subject to a condition that the librarian at institution B maintains a record that the copy was made.
The ONIX for Licensing Terms message needs to be able to handle either the generalised term or the complex form.

Further development work has been carried out and a draft format of the ONIX for Licensing Terms message with examples is now available on the EDItEUR website ONIX for Licensing Terms requires that a formal definition is provided for:

(a) Each "party" that is mentioned anywhere in the license.
(b) Each "resource" that is mentioned anywhere in the license (including resources that are derived by actions taken under the license, eg extracts made from the original licensed materials).
(c) Each "time" or "place" that is mentioned anywhere in the license.
(d) Each external "document" (paper or electronic) that is referenced anywhere in the license.
(e) Each "usage" that is referenced anywhere in the license.

In each definition, a "label" is assigned that must be unique within the License Terms document, and this label is used elsewhere in the XML to refer to the entity that has been defined.

The definitions are crucial to the ONIX Licensing Terms structure, and are likely to be the largest section of a License Terms XML document.

The key elements in the ONIX Publisher Licensing Terms format carry controlled values which will be managed through a structured dictionary – the ONIX Licensing Terms Dictionary.

Next steps
However, in addition to the technical work remaining, there are still several practical and political issues to be dealt with:

  • Publishers, especially small and medium sized ones, will need tools and services to help them produce the XML representations of their licences
  • Integrated library systems will need to implement the standards in the electronic resource management systems that they are developing.
  • Some librarians are concerned that the development of licensing messages represents "the thin end of the wedge" in terms of introducing DRM enforcement technology into the relationship among publishers, libraries and library users. (In fact, the fuller compliance to licensing terms that this work will facilitate makes the implementation of technical protection measures even less likely or attractive to publishers).
  • There are concerns that the precision required to automate the exchange of licences could remove deliberate ambiguity in a licence that is sometimes key to the successful conclusion of negotiations.
  • There are issues about whether the paper or the electronic version will be the canonical licence (and where liability lies if a system misinterprets a licence term).

A new Joint Working Party of ERMI, EDItEUR, NISO and PLS is now being set up to carry the work forward, further develop, pilot and promote the messages. The organisations forming the new joint working party and representing libraries, publishers and standards bodies are optimistic that by pooling resources and working collaboratively, these issues can be sensibly discussed and dealt with. Any readers interested in learning more or becoming involved in piloting the messages are invited to contact the author.


About the author: Brian Green is Manager of Book Industry Communications, the UK book and serial industry standards body set up in 1991 by The Publishers Association, The Booksellers Association, The Library Association (now CILIP) and The British Library. He also manages EDItEUR, the international umbrella body for book and serials standards. Brian is active on many international standards committees and chairs ISO TC46 SC9, the ISO subcommittee concerned with identifiers and metadata. Brian can be contacted at

Status: first posted 28/09/05; licensed under Creative Commons