National Library Home > News & Events
Gateways

Trove: The Library’s New Discovery Service

Screenshot of the Trove beta service.

Screenshot of the Trove beta service

In the past 12 years, the National Library of Australia has developed eight ‘national online discovery services’—free services to assist researchers to discover information resources held in Australian collections:

  • Register of Australian Archives and Manuscripts (1997)
  • Picture Australia (1999)
  • Music Australia (2004)
  • Australia Dancing (2004)
  • The search service for the PANDORA Web Archive (2005)
  • The ARROW Discovery Service (2005)
  • Libraries Australia (2006)
  • Australian Newspapers Beta (2008).

In 2007, the Library decided that combining these into one service would be in the interests of both the users and the Library. This led the Library to commence the ‘Single Business Discovery Project’ in September 2008.

The Library has now given the new service a name—Trove, as in ‘treasure-trove’, being a collection of valuable or delightful things. Trove is derived from the French verb trouver, meaning to find or discover. Thus, the name encapsulates the concepts of a collection, of treasured and valuable collection items and of the process of discovery.

The potential benefit of Trove to the end user can be illustrated by the following example. Suppose a scholar is researching the life and works of Ethel Turner, author of Seven Little Australians. Through a single search in Trove, the user will be able to access:

  • books by and about Ethel Turner, with information on the location of those books in Australian libraries
  • easy access to the full content of books, where they are out of copyright
  • articles, conference papers, theses and other research dealing with Ethel Turner, including content from university open access repositories and articles in e-journals (and in the latter case, with support for navigation to the full content where the scholar’s library has a subscription to a product containing the e-journal)
  • pictures of Ethel Turner held by libraries, museums and archives, including digitised pictures and information about the location of pictures not yet digitised
  • archived websites that refer to Ethel Turner
  • information about archival papers, letters, diaries and other records relating to Ethel Turner
  • biographies of Ethel Turner from sources such as the Australian Women’s Register, the Dictionary of Australian Biography Online and Wikipedia.

In the above example, the search of Trove retrieves multiple result sets that are presented in ‘collection views’. These views recognise the unique requirements of different types of information resources, and Trove applies its own relevance-ranking algorithm and its own facets to narrow the search results.

Many of these views will be based on the format of the information resource, for example, ‘Pictures and Photos’. Others may reflect a topic approach, such as ‘About People and Organisations’. In the future, some of these views may reflect disciplines that are represented in multiple resource formats, such as ‘Music’. And some may reflect a wider contextual perspective, such as ‘Research’. The number of collection views, and the boundaries between them, are likely to change as the scope of the service develops, and in response to feedback from users.

In late 2008, the project team assembled a focus group of potential end users to participate in a ‘card-sorting’. The aim of the exercise was to give the team a better understanding of the format categories that were understood by users and of the terminology which would make most sense to users in describing those categories. The report of this exercise was published on the Library Labs site.

Following this exercise, the project team developed the prototype based on the following eight collection views:

  • Books, Journals, Magazines, Articles …
  • Pictures and Photos
  • Australian Newspapers (1803–1954)
  • Music, Sound and Video
  • Maps
  • Archived Websites (1996–now)
  • Diaries, Letters, Archives …
  • About People and Organisations.

The last of these views, ‘About People and Organisations’, reflects the Library’s objective to support an entry point to Australian web-based biographical services. The national authority file, supported by Libraries Australia, is being used as a starting point for the program. The program will create a virtual web page for each person or organisation. This page will cite works by and about the person concerned, and will support navigation to biographical information contained in other web-based services.

In August 2009, the following biographical data had been harvested:

  • Australian Name Authority File
  • Biographical data in Music Australia and Australia Dancing
  • Australian Women’s Register.

Many other potential contributors have been identified, including the Australian Dictionary of Biography Online, the Dictionary of Australian Artists Online and Bright Sparcs (which documents Australian scientists).

The project team released a prototype of Trove in May 2009. The prototype service included a prominent ‘feedback’ box and, over a three-month period, many comments and suggestions were received.

The production release of Trove is planned for the early November 2009. The Library then plans to extend Trove by making a number of general improvements, such as an advanced search for each view, RSS feeds, enhanced sorting of results, more external targets, full text from more sources and bug fixes). We also intend to develop the Australian Newspapers collection view to meet priority enhancements identified during the Australian Newspapers Beta phase.

During 2010, the Library will commence the decommissioning of other legacy services, commencing with Picture Australia. This decommissioning will proceed only when Trove has developed features that match those of the legacy service, in terms of both the user interface and functional features, including the provision of trails and advanced searching.

As part of the Reimagining Libraries program, the Library will work with state and territory libraries to identify additional content for Trove, as well as opportunities for these libraries to build their own discovery services by leveraging off the data collection and its application-programming interfaces.

This ambitious attempt to construct a national ‘information portal’ is approaching fruition. While there are many examples of worldwide portals with extensive coverage, the National Library is not aware of any others that include content from such a variety of sources, including library, archive and museum collections, university repositories, digitised newspapers, web archives and online biographical databases.

Warwick Cathro
Assistant Director-General, Resource Sharing and Innovation