Preservation of Electronic Records:
New Knowledge and Decision-making Symposium 2003,
Ottawa, September 15-18, 2003
Bonnie Woelk, Tim Au Yeung, Maureen Hyland, University of Calgary
One of the most pressing needs facing archives and libraries is the preservation of digital records and other objects. This past September, the Canadian Conservation Institute, the Library and Archives Canada, and the Canadian Heritage Information Network hosted a symposium on this topic. Participants put forward and struggled with the formulation of solutions to a number of digitization issues that are of concern to the archives and library community. In particular, the discussion focused on the following specific topics in relation to electronic records and objects: authentication; the development of preservation strategies and standards; and preservation of archival records, audio-visual materials, CD ROMs, just-in-time websites and digital products.
In a paper entitled "A strategy for archiving websites at the Library and Archives Canada," Senior Policy Officers Susan Haigh and Roselyn Lillenilit described activities at the Library and Archives Canada aimed towards the efficient and effective provision of long-term, on-going access to significant Canadian government websites. This began with a pilot project in 2003 to explore possibilities for the harvesting of Canadian websites. Although a large number of technical obstacles arose, the pilot resulted in a short list of sites suitable for capture, and permission was obtained from some website owners for on-going capture.
Haigh and Lillenilit also reported that Library and Archives Canada is a member of the Internet Preservation Consortium, an international organization, the goal of which is to explore ways of implementing the effective harvesting of Internet content. Priorities of this group include the development of a new "smart crawler," exploration of the various national legislative needs, and solutions to a number of challenging technical concerns.
Another member of this consortium, the British Library, was represented at the symposium by Deborah Woodyard in "Digital preservation strategy at the British Library: application of the Preservation management of digital materials handbook". Woodyard provided a practical example of how the handbook can be used to provide a "digital library infrastructure" in a library setting in order to develop coherent digital preservation strategies and best practices. Recommendations were extracted from the handbook and applied to current British Library practices regarding the entire "life-cycle" of digital resources. Progress was analyzed and assessed, information was passed on to management staff, and implementation was begun where feasible and appropriate. Both the strengths and the weaknesses of the handbook were discussed.
Technical issues regarding the preservation of electronic records and digital objects were also considered at the symposium. Peter Adelstein ("Preservation of Electronic Records - Status of ISO Standards") addressed the issue of ISO standards with particular emphasis on the durability of CD media. In giving the sense that current work on the standards has stalled, Adelstein left the door open to continued work. In contrast, Basil Manns presented work done by the Library of Congress ("Longevity of CD Media - Research at the Library of Congress") to actively test the continued accessibility of commercial CD-ROMs in the Library of Congress collection. The longitudinal study undertaken provides a good idea of the kinds of activities that need to be undertaken to validate and verify beliefs about the durability of carriers for electronic records and objects.
Media does fail, however, and so the symposium provided examples of "digital archaeology" (as Seamus Ross terms it) to recover data and actively preserve it for the future. Hannah Frost ("Waiting to Happen: Lessons from Preserving Disaster-afflicted Electronic Media in an Archival Collection") presented an example of recovering media of an already neglected collection after water damage. While Frost's case study represented a collection left unattended even before the need for recovery, Michael Olson ("Robert Creeley's Computer Files: A Collaborative Approach to Link Access with Preservation") gave an example of working with an electronic records creator to actively archive the records and ensure their accessibility for the future.
A number of speakers addressed the issue of ensuring that electronic records are able to serve as trustworthy evidence and fulfill their purpose of transmitting reliable information over time. Electronic records can be so easily manipulated and are so fragile; therefore, their authenticity is constantly at risk, especially when they are transmitted through time and space and when they are migrated from an obsolete system to a new one.
Luciana Duranti ("Authenticity Requirements for Electronic Records") reported on the findings of the first phase of the InterPARES project (International Research on Permanent Authentic Records in Electronic Systems). The project has developed two sets of conceptual requirements for assessing the authenticity of administrative records generated or maintained in document management systems or databases.
The first set, called "benchmark requirements", supports the presumption that electronic records are authentic while they are in the custody of their creator. These requirements identify the information that establishes an electronic record's identity, i.e. its uniqueness from other records; demonstrates its integrity, i.e. it is intact and uncorrupted; and identify the types of procedures that govern the record's creation, handling and maintenance in such a way that the record can be presumed authentic. The requirements also address access privileges, loss or corruption of records, media deterioration and technological change, authentication of records, and the transfer of metadata when records are transferred out of the recordkeeping system.
The second set, called "baseline requirements", supports the production of authentic copies of electronic records that have been transferred to the custody of an archives/preserver. These requirements include procedures and systems used to transfer records to the archival institution and to maintain and reproduce them in a way that adequately guarantees their identity and integrity. These include ensuring that the content of records is unchanged, that reproduction is documented, and that records are transferred directly from the creator to the preserver, as well as security and control procedures.
Duranti's paper set the stage for discussion about the preservation of digital records with the implicit reminder that there is no point in preservation if the records are not authentic, i.e. if they cannot serve as trustworthy evidence. Terry Eastwood ("Appraising Digital Materials for Preservation as Cultural Heritage") assumed the fundamental necessity for authenticity and asserted the central role that appraisal plays in the preservation of digital materials. He outlined lessons from archivists' experience of appraising electronic records that can be applied to digital cultural heritage preservation in general.
Appraisal involves estimating or judging the worthiness of records or other cultural materials of continued preservation. Eastwood argued for a framework of policies and procedures to guide this selection function. He identified four activities in appraisal: 1) compiling and analyzing information about the digital object(s), which is essential to understanding the value of records because their context within a body of records created by a particular creator gives them their meaning; 2) assessing the capacity of records to serve the continuing interests of their creator and society, and digital objects to serve as expression of cultural heritage; 3) determining the feasibility of preserving the records so that their authenticity is maintained given the current and future capabilities to preserve them; and 4) based on the foregoing, making the appraisal decision and carrying out the disposition by first setting out the terms and conditions for the transfer of records so that their authenticity is preserved.
Bruce Walton took up this last issue and addressed the question of how to effectively preserve digital records once they are in the custody of the preserver so that their authenticity is maintained. He discussed the preservation methods currently available and their strengths and weaknesses: leaving the digital components of the record as received and concentrating on maintaining the technology needed to reproduce it; leaving the digital components as received and emulating the old technology on new technology; continually updating the logical format of the records so they can be reproduced using current technology; and converting records to standard non-proprietary formats. Based on the experience at the Library and Archives Canada, Walton recommended the conversion of records to a standard format such as XML and reliance on the procedural and technological context of the recordkeeping system rather than any individual qualities of the records in the system for record authenticity.
The symposium's organizers wisely included papers about current projects that are applying current research to practice. Howard Quenault presented a case study describing current efforts to preserve digital records and archives at the Public Record Office Victoria in Australia. The Victorian Electronic Records Strategy (VERS) was developed in the late 1990s to provide a framework of standards, guidance and implementation projects aimed at preserving authentic government electronic records. The VERS solution consists of converting records to a long-term non-proprietary format, encapsulating records with a set of metadata using XML, and attaching a public-key digital signature to ensure record authenticity. The VERS strategy was implemented in a major government department as a pilot project (1999-2002) that captures records at creation, manages them in a VERS compliant recordkeeping system and exports them to the Public Records Office when required. The VERS program will be implemented in the other major departments of the government in 2002-04.
The papers of Quenault, Duranti, Eastwood, and Walton gave a good overview of the theoretical and methodological aspects of authentic electronic records preservation from the creation, maintenance, and appraisal of records to their transfer to the custody of an archives/preserver and their subsequent preservation over time.
While it is clear from the papers presented at the symposium that much activity is occurring because the need is very pressing, the guidance for institutions on issues of digital preservation is still being formulated and comprehensive solutions remain to be developed. If anything, the Preservation of Electronic Records symposium represents the beginnings of a dialogue in Canada on the need for concerted, deliberate preservation of electronic records and digital objects.