ALEG

Stage 1 Report - 15 May 2000

Draft 4, 15 May, KF
Introduction

The ALEG project is a unification of the nation's major and most authoritative creative literature resources:

The project aims to create the most comprehensive and authoritative web based resource for scholars of Australian creative literature, building on the extensive coverage of the above individual resources by providing: This document describes the outcomes of Stage 1 of the technical implementation of the ALEG project. The deliverables of each stage of the technical implementation are as follows:

SCOPE & FUNCTIONAL REQUIREMENTS

Goals of the project

  1. unify the key component ALEG resources under a single hardware/ software platform.

    The component resources are currently managed by a variety of hardware/software systems. The AUSTLIT database uses obsolete infrastructure which needs to be replaced. Some others are not accessible from the web, some cannot be easily linked and cross referenced.

    [AUSTLIT was developed by Library staff in 1988. It was based on the AWA URICA product, for which the Library obtained source code under the terms of their agreement with AWA. Although this platform has served AUSTLIT well in terms of functionality, performance and reliability, URICA and the environment it uses ("PICK") are discontinued, non-mainstream products and support issues will soon arise. Additionally, the goals of the ALEG project require major rearchitecting of the existing AUSTLIT database and user interfaces, as described below. It is not possible to economically undertake this work within the URICA/PICK emulation infrastructure.]

  2. design and implement record structures to support powerful searching and flexible representation of more information and relationships across ALEG's component resources.

    Over a decade of use of AUSTLIT has generated requirements which cannot be met by the existing AUSTLIT database structures. The ALEG project will result in an expansion in the types of information being stored, most significantly due to the incorporation of the Bibliography of Australian Literature Project (BALP).

    Whilst the search capabilities of the current AUSTLIT system are probably adequate for casual users, search technology and hence expectations have increased greatly over the last decade, and the ALEG database will need to support at least the flexiblity of the CD version of AUSTLIT, but delivered over the Web.

    The new system will also support ranking of resources based on citation and subjective importance, although users will be able to rank search results by other attributes such as publication date, title, author.

    Among the existing AUSTLIT data design areas to be addressed are:

  3. development of the ALEG Web interface

    The AUSTLIT web interface was implemented in 1997. Although it is simple and functional, 3 years is an eternity on the web. Enhancements to the web site have been identified but left on hold pending the ALEG redevelopment. Specific areas to be addressed are:

  4. support easy-to-use and distributed database maintenance

    The current AUSTLIT data entry system is both very limited in capability and requires extremely skilled and specialised knowledge to operate.

    The redeveloped system will allow more efficient data maintenance by both specialist data entry staff and subject-matter experts. It will support:

    The emphasis will be on productivity, ease of use and support for data integrity/consistency goals of ALEG.

  5. support a Z39.50 query interface

    ALEG will implement a Z39.50 target, allowing searching of its publically available data from Z39.50 clients and gateways. The Z39.50 "Bath profile" specification will guide the implementation of the target.

  6. support MARC output

    The Z39.50 target requires MARC output capability. Hence, regardless of whether ALEG information is partially or completely stored in a MARC representation, it needs to be mappable to MARC.

  7. support XML output and customisable data extracts for specialist users

    The many specialist providers of information to ALEG will require data extracts of selected material in specific formats. XML is becoming a widely and successfully used data interchange format, and regardless of whether ALEG information is partially or completely stored in an XML representation, it needs to be mappable to an XML Document Type Definition (DTD) or Schema. Having an extract in XML format will facilitate the development of XML Style sheets for the transformation of a specific XML format into a form required by specialist users.

  8. support an enhanced thesaurus structure

    The current AUSTLIT system supports thesaurus and non-thesaurus terms, but a non-technical goal of ALEG is the review of thesaurus strategies, possibily resulting in a richer thesaurus structure (broader and narrower, "see also", "use") and possible automatic use of thesaurus information to automatically or semi-automatically aid searchers (eg, augmenting a search specifying 'color' with '(color or colour)', suggesting broader or narrower terms based on the volume of search results).

  9. support access to holdings information

    The current AUSTLIT system offers users a standard Inter Library Loan form for some works physically held by ADFA.

    The ALEG system will allow the user to view holdings information for all works, ranked where possible by estimated physical proximity to the user. Exactly how this will be done requires further investigation (see "Open Issues" below).

  10. link to manuscript & archive information

    The NLA's Register of Australian Archives and Manuscripts (RAAM) database contains information about Australian manuscript collections. It is desirable that ALEG users can be linked to this resource where appropriate, either by including a specific link stored within ALEG against certain resources or just by providing a hyperlink to the RAAM search function (another "Open Issue").

Non-goals: what we are not attempting to do / what is out of scope

  1. stored/recurrent search ("SDI")

    SDI searching facilities will not be provided in the initial implementation

  2. non Web (eg telnet, stand-alone software) client

    The primary access mechanism will be the Web. The ALEG system will not be accessible by telnet or any special ALEG-specific client. The current AUSTLIT CD will be discontinued. However, the ALEG Z39.50 interface (web gateways or Z39.50 clients) will provide another access mechanism. As well, consortia members will be able to selectively extract the information they want from the ALEG and make it available in alternate forms (web site, printed document, etc).

  3. spell check

    The data maintenance facility will not attempt spell checking. (However, compliance with thesaurus terms and other authority files will be checked where appropriate.)

Open Issues for Stage 2

Preparation of the Stage 1 report has highlighted some areas which will require particular attention during Stage 2 (Stage 2 is the development of a detailed technical design and implementation plan):

  1. implementation infrastructure

    There are 2 'existing infrastructure' platforms which are contenders to be used to implement the ALEG platform:

    There is also existing infrastructure from a related Australian project (the "Online Heritage Resource Manager" produced by the Australian Science and Technology Heritage Centre) which is perhaps more closely aligned to the complex data representation and relationship goals of ALEG than a text based or library catalog/MARC based system.

    Before a decision can be made on which infrastructure platform should be used, or what can be reused and what can be built from high level components, more work needs to be done on the exact data management and reporting requirements of ALEG and how these are expected to evolve in the future. This work will begin immediately and be completed during Stage 2.

  2. access policy

    It is not yet clear what information will be freely available and what will require subscription or per-view payment. However, what is known is that the system will need to support access policies which are flexible and easily modifiable over time.

  3. distributed management and data maintenance workflow

    ALEG represents a move from smaller, independently managed resources to a larger, distributed, cooperative managed resource. The advantages of a more comprehensive and cohesive collection are many, but this evolution will require carefully considered system support for local work practices, distributed centres of knowledge and control, and for creating an effective work environment for the "community" of contributors.

  4. support access to holdings information

    The current AUSTLIT system offers users a standard Inter Library Loan form for some works physically held by ADFA.

    The ALEG system will allow the user to view holdings information for all works, ranked where possible by estimated physical proximity to the user. Exactly how this is best done requires further investigation (see "Open Issues" below).

  5. link to holdings information and manuscript information

    It is desirable that ALEG provides authoritative links to information where possible. Holdings information could be very useful for ALEG users wanting to access the resource. However, although NLA's Kinetica is a good source for much holdings data, there are open issues on how best to access on this information using Kinetica for a specific resource, and open issues on whether libraries which do not store complete holdings information on Kinetica should be accessed as well or in place of Kinetica. Maybe for some ALEG resources, holdings information will be recorded within ALEG.

    For manuscript collection information, ALEG could selectively "point" the user at RAAMS, perhaps with a suggested search term (stored within ALEG), or perhaps provide additional manuscript collection information stored within ALEG.


Home > Stage Reports
Kent Fitch
k.fitch@adfa.edu.au
15 May 2000