PANDORA Logical Data Model, Version 2, November 1997Introduction
Key Data Requirements
Key Interfacing Requirements
The PANDORA Project (Preserving and Accessing Networked Documentary Resources of Australia) was set up in the second half of 1996 to progress the initiative of the National Library of Australia to ensure long term access to significant Australian on-line publications.
The aims of the project are to:
- identify the functional requirements for a management and preservation system.
- identify, test and evaluate a range of technical methods, standards and products involved in the preservation process, including capture, cataloguing and archiving.
- estimate the financial, equipment and staff resources required for the ongoing management of publications selected for national preservation.
- develop recommendations and strategies for the long term preservation of and access to publications selected for national preservation, including considerations of data refreshment, reformatting and migration to counter obsolescence or loss.
- develop a proposal for a national approach to the long-term preservation of these publications.
Prior to establishing the PANDORA Project, the National Library had set up the Selection Committee on On-line Australian Publications (SCOAP) to develop guidelines for the selection of titles to be preserved by the National Library. It had also set up the Australian Electronic Unit to implement the recommendations of SCOAP and to identify, select and catalogue significant Australian titles. A small proof of concept archive was established to test technical capability and business principles.
An archive management facility to support a large scale archive and the procedures and operations involved was now required. Four main components in the development of such a facility have been identified:
- the Business Description (completed)
- the Logical Data Model (this document)
- the Technical Infrastructure Document (in preparation)
- product selection (to be informed by the three previous components)
The Business Description is crucial to the process of developing the facility and should be read in conjunction with the Logical Data Model.
There is significant overlap between the data requirements for a digital archive such as PANDORA and for a digital library. The primary difference is the need to ensure the integrity and long term accessibility of the digital objects selected for inclusion in the archive. The creation of digital archives is a relatively new activity and it will be important for the PANDORA Project to participate in and assist with the development of national and international data standards in the area of digital archiving. The PANDORA logical data model is based on the experience which has been gained so far from the proof of concept archive and is also informed by the following related initiatives:
- Research Library Group. Task Force on Archiving of Digital Information. Preserving Digital Information: Final Report and Recommendations, 20 May 1996
- Consultative Committee for Space Data Systems. Reference Model for an Open Archival Information System (OAIS) (CCSDS 650.0-W-1White Book, Issue 1, 10 April 1997). This is an ISO initiative. The ISO Archiving Standards home page is at http://ssdoo.gsfc.nasa.gov/nost/isoas/. The latest version of the Reference Model is at ftp://nssdc.gsfc.nasa.gov/pub/sfdu/isoas/us07/CCSDS-650.0-W-1.m51.
PANDORA falls into the category of an active, information preserving, permanent archive in that it needs to:
- take in new "versions" of objects in the archive to reflect changes to the original copy over time;
- preserve both data and information interpreting what the data means; and
- deal with issues related to ongoing access as technologies evolve.
The archive is not a collection in its own right, but is intended to act as a digital object server for the Library's Online Public Access Catalogue (OPAC). Ultimately, it will form part of an Australian Digital Archive accessible via the National Bibliographic Database (NBD). This will be achieved through the collaborative efforts of a range of partners committed to ensuring the integrity and long-term accessibility of significant Australian online publications falling within their sphere of responsibility.
The priority for the Library is to make publications selected for the PANDORA archive visible to users at title level. Having identified a title, they can then browse through the Library's holdings of copies of each title; or exploit links to the publisher's site if it is still active. The content will be completely opaque to the search engine. The access function will only process queries based on metadata for each title. Opportunities will exist via a range of other resource discovery services to index the content of PANDORA archive objects for access at a more granular level. However, this falls outside the scope of the PANDORA Project.
Involvement of the PANDORA archive management facility with a title begins at the identification stage. Preliminary checks are made to assess its suitability for archiving and the results are recorded. Additional metadata is input once the selection decision has been made and at subsequent stages, as the publication is made ready for archiving. Unique identifiers are allocated. The title is technically evaluated. The publisher is contacted and the Gathering Regime, file structures and access conditions negotiated. Information is recorded about any special software or conversion requirements and the item is fully catalogued.
As each copy of the object is added to the archive, additional metadata is recorded about the copy. The object itself may also be modified after it is captured, i.e., to de-activate payment functions, remove duplicated content, delete unwanted or extraneous parts or de-activate links. These changes need to be recorded to ensure users can trace the provenance of the item and be assured of its authenticity. From time to time, changes to the status of the title or to the format in which copies are stored and made accessible via the archive may also need to be recorded.
Access to the PANDORA digital object server will be free of charge to the end user and, as far as possible, both internal and external users will have unrestricted access. However, there will be some occasions when it will be necessary to restrict access, for instance, when unlimited access would jeopardise a publisher's vested commercial interest or when there is a temporary legal embargo.
The archive management facility will be required to manage circumstances where access may be provided to users within the Library building but not those outside it. It must also be able to inhibit access to designated publications (or parts of publications) before a specified date, but provide access after the date has passed.
All of the functions of the archive will require the production of management reports. These will draw from logs at both system and application level to produce analysed information as required for decision support purposes. The reports will need to be run on both a prescheduled and an ad hoc basis.
The Library has an Integrated Library Management System (ILMS). This is used to automate the selection, acquisition, accessioning, cataloguing and storage of materials in the Library's collections; and to make these materials accessible to end users. The functions needing to be supported to process online publications for inclusion in the Library's collections are a superset of those needing to be supported for physical format publications. Descriptive metadata for each title created at the identification stage forms the basis for the full catalogue record made accessible to users on the Library's ILMS and on the National Bibliographic Database. A record will also need to be kept in the ILMS Acquisitions Module when there is a charge involved in acquiring a title for the PANDORA archive or making it accessible to users..
During the negotiation of agreements with publishers, a significant amount of correspondence may be generated that will need to be stored in the Library's Record Management System (TRIM Enterprise). The agreement with the publishers will define procedures for capturing digital objects from the publishers' site. Publishers may create their own descriptive metadata for titles selected for inclusion in the PANDORA archive and there may be opportunities to use this metadata as the basis for the Library's record. The PANDORA archive management facility may also capture descriptive metadata from other repositories as part of the process of identifying candidate titles for inclusion in the archive.
Objects in the archive may be accessed from the NLA OPAC. They may also be accessed from the NBD or any other resource discovery service that has obtained a copy of the catalogue record or indexed the content of archive objects for their own business purposes. Access will be effected via a unique URL for each title. To ensure the permanency of the URL as the physical location of the archive copy changes, the URL will be created as a PURL.
A PURL is a Persistent Uniform Resource Locator. Functionally, a PURL is a URL. However, instead of pointing directly to the location of an Internet resource, a PURL points to an intermediate resolution service. The PURL resolution service associates the PURL with the actual URL and returns that URL to the client. The client can then complete the URL transaction in the normal fashion. In Web parlance, this is a standard HTTP redirect.
These relationships are illustrated in the diagram below:
The entity-relationship diagram identifies the logical entities needing to be supported by the PANDORA archive management facility and the relationships between them.
The diagrammatic representation used follows Information Engineering conventions: