Paper presented by Margaret Phillips, Manager, Electronic Unit at the Information 'Online & On Disc 97' conference, 21-23 January 1997 Sydney Convention & Exhibition Centre Sydney.
I want to tell you about the work we're doing at the National Library of Australia towards ensuring long term access to online Australian publications. The written version published in the conference proceedings outlines what we'd done up to the end of last October. In this presentation I'll summarise the written paper and deal in more detail with what's happened since.
On the screen now there's an outline of my presentation. After a brief introduction, I'll list the Library's initiatives and then go on to describe them in more detail.
I scarcely need to tell this audience that the role of libraries in the era of electronic publications has been the subject of much debate. At a time when library budgets and staff numbers are declining, the prospect of coming to terms with a new, resource intensive format is an enormous challenge. Nevertheless, the National Library considers that its responsibility 'to maintain and develop ... a comprehensive collection of library material relating to Australia and the Australian people'[1] embraces electronic publications. We're arguing that they are new business and are requesting additional funds to manage them.
As you all know, there are two types of electronic publications: physical format items such as CD-ROMs and floppy disks, and online or networked publications. While the National Library has also been working towards developing a national collection of physical format titles[2], this paper deals with what we're doing in relation to online publications.
The National Library is willing to take a leadership role in developing a collaborative approach towards guaranteeing long term access to Australian online publications. During 1996, the Library undertook a number of initiatives in this direction:
I'll elaborate on each of these shortly but first let me tell you about some other outcomes
I'll now discuss each of the Library's initiatives in more detail.
In January 1996, the Selection Committee on Online Australian Publications (SCOAP) had its first meeting, with representatives from Acquisitions, Cataloguing, Preservation, Information Services, ABN, Serials and IT Facilities. We considered it essential to be able to canvass a broad range of expertise within the Committee.
Its purpose was to develop guidelines for the selection of those Australian online publications which the Library would want to undertake to preserve. Our brief was to focus on what, as a National Library, we should want to be able to do. To avoid being bogged down in technical and resource issues, we were not to consider how it could be done, only what ideally should be done. This proved to be a very liberating approach and helped us to avoid the circular discussion that had been a feature of previous meetings on the subject.
As a starting point we took the Library's Collection Development Policy for print Australiana. In principle, we considered that the format of an item should make no difference to whether or not it's preserved. The content of the item should be paramount. Where types of publications had no exact print equivalent (home pages, for example) the same principles were to be applied.
SCOAP adopted a practical approach and managers who were unused to dealing with actual materials surfed the Net and became familiar with online publications at first hand. This was important. Only then could we make realistic and informed decisions about them. The main item of business at each meeting was an assessment of a particular category of online publications: conference proceedings, literary works, government publications, e-journals, home pages, to name some that we dealt with in the first few months. A week before each meeting, Committee members would be provided with a list of about ten titles and their Internet addresses. They'd locate these titles on the Internet and evaluate them, using a work sheet designed for the task. This was expensive in terms of staff time but it paid results.
From the decisions made at each meeting in relation to particular titles, principles for selection were extrapolated and gradually the selection guidelines took shape. They're now available on the Library's server[6].
From that description of proceedings, you may think it was all very easy. It wasn't. As convenor of SCOAP, I sometimes despaired that we'd ever achieve our objective. Decisions that Committee members were unhappy with came back time and again for review.
The principal issue that we grappled with was a fundamental one: what constitutes an Australian publication in the electronic environment? In print, this is usually more clear cut. For instance, in the National Library, if an item is published in Australia, has an Australian author or has at least ten per cent Australian content it's taken into the Australian collection. Does the mere fact that an electronic publication has an Australian Internet address make it an Australian publication? It depends on your definition, of course, and this is just what SCOAP was attempting to define. The Internet has broken down geographic boundaries. A journal can be edited in the U.S. one year, Australia the next, Canada the year after. This does happen in print, but not quite to the same extent. An e journal might be edited in Australia, but its principle site may be in the United States. The crucial point here is - which country will take responsibility for preserving it?
One of the handicaps we've been struggling with is the lack of a taxonomy for the various types of networked publications. In the Library environment, the ability to define and categorise information is an important part of organising and providing access to it. With the obvious exception of 'home page', the terms we have to describe online publications are derived from the print environment and sometimes there are no useful terms. For instance, what should we call a publication that refers to itself as a 'zine', a 'magazine' or a 'journal', but has no discrete issues. Text is added, often on an irregular basis. Sometimes earlier segments disappear. For want of a better term, we've been calling them 'evolving serials', but this isn't a standardised term and therefore not commonly understood even by colleagues elsewhere in the Library.
One of the assumptions underlying all of our work with online publications is that, even with wide ranging collaborative effort among libraries and archives, we won't be able to preserve the entire national output online. We must be selective.
In April of last year, the Library established the Electronic Unit. Its main tasks were to:
I've elaborated on the main issues that we encountered in our work in the Electronic Unit in my printed conference paper. In order to have time to tell you about more recent developments, I won't go into detail on these issues now. Suffice it to say that, in almost every respect, electronic publications seem more complex or more labour intensive than we've been used to with print.
The current focus of our attention is PANDORA. As you may know, there are two versions of the Pandora myth. In the earlier version there evil things in the box. We've adopted the later myth which has good things in the box, in our case, an archive.
In June, this project, Preserving and Accessing Networked Documentary Resources in Australia, was established to:
As you will have gathered by now, we're very keen on quantifying resources.
As with SCOAP, the project team developed its model through work with specific titles, subjecting them to detailed analysis in order to ascertain the kinds of factors that we'd need to take into account when archiving them.
We decided to test as wide a range of publications as we could find, both the simple and the complex. It's important that we know the full extent of the resources that will be required to create and manage an archive of nationally significant titles. To attempt only the easy ones wouldn't give us the complete picture.
The deliberations of the project team have focused on three main points:
I'll elaborate on each of these three areas.
Firstly, The nature of the publications and the capture strategies.
The wide variety of publications we intend to include will mean the need to respond to the requirements of a number of different publishing patterns. Monographs and serials following the familiar print pattern should be relatively easy to capture. Other types of publications will require more ingenuity. Home pages can change frequently but at irregular intervals. The best solution here may be a regular snapshot at a specified interval, perhaps one month, perhaps once in 12 months depending on the nature of the publication.
Where an 'evolving serial' continually cumulates, we should be able to overwrite the version already in the archive with the new version. The trickiest scenario of all is the 'evolving serial' which both adds to and subtracts information from the publication at any time. Close cooperation with publishers will be necessary for the successful archival management of these titles, in particular. We regard close cooperation with publishers and creators as an essential aspect of the archiving project in any case.
Secondly, the Technical and procedural issues addressed by PANDORA.
The Library's Server will be used to store archived documents in the short term. In the long term we anticipate employing some off line storage also.
As I've already mentioned, before a title is included in the project, we undertake a technical evaluation. It's a time consuming process (an average one hour per title) which looks at:
We'll indicate to users and other libraries that the National Library intends to archive a particular title to avoid duplication of effort. It's very important that this information should be readily available. In consultation with the ABN Office, we decided to use the Action Note in the NTA field of the catalogue record for this purpose. (It's the 583 field if you're more familiar with USMARC.)
We've been experimenting with capture software that will enable us to transfer publications from the Internet to the Library's Server. At the time of writing the published paper, we were looking at WebWhacker[7] and Freeloader[8]. We're currently working with the Harvest[9] software developed at the University of Colorado. Harvest was considered a good candidate for further exploration because it's highly configurable, enabling document selection based on U-R-L, file name and content, and providing hooks for external software modules.
The last of the three main areas that PANDORA has focused on is Access conditions to archived publications.
The project team has developed a service model which conceptualises how we'll provide access to the Library's archive of online publications and what will be required to implement it.
Access will be provided to a title via a link from the ELA field in the catalogue record to the item itself in the archive. (In USMARC, it's the 856 field.) The hot link facility made available by a World Wide Web interface with the Library's OPAC will be used. The link will take the user initially to a title screen in the archive that will provide information about the publication, including the Library's holdings.
The title screen will also provide the option of going directly to the publisher's site, if this is still active. In fact, there will be links to the publisher's site, both from the catalogue record and from the title screen in the archive.
An initial 25 titles were selected and publishers approached for permission to archive. Twenty of the publishers responded enthusiastically in favour of participating in the project and this is an encouraging result. We're in the process of contacting more.
PANDORA is being funded in part by the Australian Vice Chancellors Committee Working Group on Electronic Publishing. An advisory committee comprising representatives of creators and publishers, collectors and users of online publications is being set up to guide the project and will have its first meeting here at the Conference. The Library considers it a great advantage to have representatives from the university, State library and publishing sectors involved in the project .
The last of the National Library's initiatives that I'll talk about today is the position paper, National strategy for provision of access to Australian electronic publications. It recommends a cooperative approach from libraries, archives, producers and creators towards guaranteeing long term access to electronic publications. The resource-intensive and specialised technical nature of the preservation activity makes this essential. Libraries and other institutions with a mandate to build deposit collections will have a key role but other libraries may accept responsibility too, particularly for publications of their own institution.
The paper also makes recommendations about
Comment on the paper was invited and responses were generally supportive. There was strong agreement on the need for legal deposit of electronic publications and for collaboration on their preservation.
There's a need for reform to legislation covering legal deposit, both at the Commonwealth and State level. The scope of publications covered must be extended to include all types of electronic publications. Institutions accepting responsibility for preserving electronic publications will need to collaborate on developing selection guidelines and enter into agreements that will ensure the nation's documentary heritage is preserved for long term access. Deposit libraries need to work together with publishers and creators to develop levels of access that are satisfactory to all parties. Fair dealing principles and copying rights need to be worked out. Standards for the creation and use of metadata need to be established.
There's still much work to be done, particularly in the area of building the archive of nationally significant titles. There is also a long way to go in forming partnerships with other libraries, archives, creators and publishers to ensure the nation's documentary heritage in electronic form is preserved.
We've made a beginning on this. On 6 March a National Metadata Seminar, being held in conjunction with the Fourth Dublin Core Workshop, will take place at the National Library in Canberra. Anyone who is a creator of information resources on the Internet is encouraged to attend and representatives from libraries, museums, archives and government agencies are particularly welcome. Copies of the brochure are available from the Library's stand here at the Conference. The National Preservation Office too is planning a meeting of key parties towards the end of this year.
The number of electronic publications available will increase,
perhaps dramatically, over time. The number of staff available to process them
will not. We must find more and more efficient ways of handling them,
developing software to automate procedures and to keep manual intervention to a
minimum. It will be a continual challenge to keep abreast of new technologies
as they arise and to use them to our advantage. As this paper demonstrates, the
National Library is committed to ensuring long term access to online Australian
publications and we've already made significant progress towards achieving this
goal.
| About This Site | Copyright | Privacy | Accessibility | Site Map | Site Search | Content A-Z | Contact Us |