Like libraries, museums face the daunting task of preserving our cultural heritage, whilst also striving with the often conflicting need to make that cultural heritage available to us today.
Perhaps differently from libraries, interpretation plays an important part in the work of museums, where exhibitions are often designed around the telling of one or more stories about the past in order to place objects within their historical context.
Possibly even more so than in the library sector, the information revolution has had a profound effect upon the ways in which museums manage information, whether for internal use within their rich collection management systems, or externally in the content made available on web sites, in 'exhiblets', or in data exchanged between institutions or embedded within various educational resources such as CD–ROMs.
For ten years now, the international Consortium for the Computer Interchange of Museum Information (CIMI)  has been at the forefront of work to bring the information revolution to museums. Past work has included important conceptual modelling of museum collections, the creation of an early Z39.50 Profile  and a rich SGML DTD , as well as extensive testing of the Dublin Core . In 2000, this work continues with examination of diverse issues from the role of WAP phones and Personal Digital Assistants in the museum world through to examination of distributed metadata harvesting. Ten years on, the work of CIMI remains important, and its relevance to the related agendas of the Joint Information Systems Committee  (JISC) has been recognised by the JISC's becoming a CIMI member, and taking a seat on its Executive Committee.
This paper illustrates some of the past and future work of CIMI, and hopefully goes some way towards demonstrating the clear relevance of this work to many areas of the JISC's activity.
CIMI is an international organisation committed to bringing museum information to the largest possible audience. Its 25–plus members, comprising museums, systems vendors, governmental and non–governmental organisations share a common mission: to encourage an open standards–based approach to the management and delivery of digital museum information.
Since forming in 1990, CIMI has made substantial progress in researching for the museum community, standards for structuring its data and for enabling widespread search and retrieval capabilities. CIMI's work is largely carried out through collaborative demonstration projects that inform not only its members, but the wider cultural heritage community, helping them learn how complex museum information can be standardized and therefore made accessible electronically.
CIMI membership is open to any institution, organization, corporation or individual — whether for–profit or non–profit. In allowing vendors, museums, standards developers, service providers and government agencies to work together, CIMI has helped engender an atmosphere of trust and progressive thinking — essential survival rations in a relatively small market. For example, this is a rare opportunity for the vendors of heritage information management packages to collaborate outside the context of the marketplace and in doing so, the museum sector has predated by some years joint ventures such as Symbian — uniting to concentrate and assert the potential of an otherwise dispersed community.
An excellent example of CIMI's approach to working across the cultural sector to test and evaluate emerging standards on behalf of the museum community is its recent review of the Dublin Core Metadata Element Set . Involving seventeen institutions from Australia, south–east Asia, Europe and North America, the test bed ran from March 1998 and was concluded in January 2000. It had a number of outcomes of significant value to the museum community, aiming not to provide an isolated view of the world, but to establish means of allowing museums to integrate emerging best practice into their work. Building a database of Dublin Core records of museum collections, the project immediately resulted in an XML DTD used to construct the database and subsequently, 'best practice' guidelines  for museums to adopt when implementing the Dublin Core. The findings of the test bed affirmed the ability of the Dublin Core to act as a high–level resource–location tool which was useful for museum collections. The test bed established however, that Dublin Core Semantic Refinements did not enable retention of integrity of museum information at a detailed level. That these facts could be established in a well–planned and rigorously implemented evaluation, involving systems specialists, museums and other organisations in the sector, provided the community with a valuable understanding of the issues inherent in implementing Dublin Core. This understanding has been further enhanced with the establishment of the CIMI Institute. Intending to raise awareness and cultivate expertise in the community, the CIMI Institute's initial activity has been a successful series of seminars and workshops on the implementation of Dublin Core. With the assistance of additional funding from the Getty Grant Trust, these workshops have been held in Europe, Canada, the USA and Australia and are set to expand in terms of the number offered and subjects covered, in the coming two years.
Other projects have included the pioneering development of a profile for museums for use with the Z39.50 protocol for distributed access to databases , and the development of an SGML DTD  for the online delivery of rich–text information such as exhibition catalogues. These, and other projects such as the evaluation of integrated information management methodologies in museums, have brought direct benefits to its members and the wider museum community alike.
CIMI plans for the next two years are equally ambitious and plan to deliver a wide range of benefits to a growing membership. They include working with the mda  (formerly the Museum Documentation Association) to establish an XML DTD for SPECTRUM, the UK Museum Documentation Standard . Once completed in Autumn 2000, the DTD will be evaluated by CIMI members in a further test bed. Benefits of this work will include the elimination of costly and time–consuming bottlenecks experienced by museums and vendors alike when migrating data between systems. An XML DTD for museum information would also permit easier integration of information across diverse systems in use within single organisations; e.g. between specialist collections management applications and back–end web management databases.
Plans are also in progress to exploit the burgeoning use of mobile communications not only in Europe, but via test bed projects based also in Japan and North America. How can museums 'push' information via WAP and other protocols to handheld devices? How will the cultural sector need to prepare itself in order to capture this rapidly expanding market — and what will it mean for the scope and nature of the information to be delivered? These and other questions will be addressed in CIMI's Handscape project, which will work with leading industry bodies to inform the museum community.
A further project will include the evaluation of scalable solutions for the interoperability of metadata from different domains, in association with the ABC initiative's  experimentation with metadata harvesting. These and a range of other, related projects will work together under the umbrella of MIDIIS (Museum Initiative for Digital Information Interchange Standards), CIMI's longer–term strategy aimed at increasing the community's understanding of the nature of the museum information universe and how the types of information contained therein can be shared, re–packaged, and re–used for a variety of applications.
Finally, the CIMI website is also undergoing a major review and will be relaunched in October 2000, not only with additional member and staff services, but implementing the standards which CIMI has been active in promoting. Together with a new series of 'industry innovation' briefings for the cultural information management sector, CIMI will continue to provide a trusted and authoritative source of specialist advice and knowledge.
The clear relevance of CIMI's work to the UK university museums within JISC's purview aside, much of CIMI's effort is also applicable to those aspects of tertiary education more traditionally associated with JISC; our developing electronic 'libraries'.
CIMI's work, in essence, focusses upon exploring approaches to the management of information in a digital environment. Behind the community language, and the realia themselves, the issues of packaging, formulating, exchanging and managing this information are remarkably similar across all memory institutions; and beyond. CIMI's Handscape project, for example, aims to explore the practicalities of extracting rich — often lengthy and learned — descriptions from existing content repositories and packaging these for delivery over a multitude of channels and to a variety of audiences. That the content repository is a museum collection management system, the resource being described a physical object from the collection, and the primary channels of interest those relating to portable devices, in no way detracts from the importance of the underlying exploration; one of content repurposing with minimal human intervention.
Elsewhere, early CIMI work has also had an impact upon JISC activities. The Arts & Humanities Data Service (AHDS) , for example, was a CIMI member prior to JISC's joining, and there was valuable cross-fertilisation between AHDS' catalogue development efforts and the CIMI Dublin Core Testbed.
The Bath Profile , too, which currently forms an important foundation to the DNER , learned a great deal from the development of CIMI's Z39.50 Profile , and the two are likely to become ever more closely linked in the future.
Cultural heritage content is becoming increasingly visible within tertiary education, whether in the form of excavation and survey archives served from the Archaeology Data Service (ADS)  or within rich image collections such as AMICO  and SCRAN , both recently made available to the community as part of the DNER's growing body of multimedia content . The work of CIMI and others ensures that such content is available, usable, liable to be accessible in the long term, and capable of effective interoperation with other resources. JISC's involvement will mean that the needs and experiences of the non-museum sector are also represented, and will assist in the dissemination of CIMI's work to a wider audience.
CIMI services are provided by 'Team CIMI'; consultants based in the cultural heritage and information management sectors. For further information about CIMI in Europe, contact Alice Grant at: firstname.lastname@example.org.
c/o Academic Services: Libraries
University of Hull
Alice Grant Consulting