How to utilise Trove in your organisation

Author: 
Debbie Campbell
Publication date: 
Tuesday, 1 June, 2010
Abstract: 

This paper introduces Trove, the National Library’s newest discovery service, and provides information to understand and market Trove, as well as describing mechanisms for exploiting it.  There are four key drivers which make Trove an important development for collecting agencies around Australia:

  • the commitment of the National Library to Libraries Australia and Trove, exemplified by long term dedicated marketing support, an ongoing development plan, and the Library’s strategic objective to “meet our users’ needs for rapid and easy access to our collections and other information resources”;
  • the National Library’s successful track record in the provision and maintenance of collaborative, innovative predecessor web-based services such as Picture Australia, PANDORA, Australian Newspapers 1803-1954 and Australian Research Online;
  • the decision to incorporate as much digital content as possible into Trove and its precursors, including content beyond Australian shores and content created the public;
  • the National Library’s availability to support mechanisms for technical integration between Trove and library Online Public Access Catalogues (OPACs). The paper outlines several options for this purpose.

Introduction

This paper introduces Trove, the National Library‟s newest discovery service, and provides information to understand and market Trove, as well as describing mechanisms for using it. Trove may affect reference services and discovery platforms but not existing data contribution arrangements to National Library discovery services.

There are four key drivers which make Trove an important development for collecting organisations around Australia:

  • the commitment of the National Library to Libraries Australia and Trove, exemplified by long term dedicated marketing support, an ongoing development plan, and the Library‟s strategic objective to “meet our users’ needs for rapid and easy access to our collections and other information resources”;
  • the National Library‟s successful track record in the provision and maintenance of collaborative, innovative predecessor web-based services such as Picture Australia, PANDORA, Australian Newspapers 1803-1954 and Australian Research Online;
  • the decision to incorporate as much digital content as possible into Trove and its precursors, including content beyond Australian shores and content created the public;
  • the National Library‟s availability to support mechanisms for technical integration between Trove and public portals. The paper outlines several options for this purpose.

Part 1 of the paper provides information and messages which can be adapted by marketing and reference services staff when raising awareness of and talking about Trove. Part 2 of the paper outlines technical details for exploiting the Trove service in local discovery platforms.

PART 1 THE SCOPE OF TROVE

Background

What is Trove?

Trove is an exciting, revolutionary and free search engine for all Australians. You can search a large and unrivalled repository of Australian material at the click of a button.

Why was Trove created?

The National Library of Australia wanted to make finding and getting information easier for all Australians. Trove provides a quick, easy and powerful way to search. Finding information is now an effortless and enjoyable experience.

Is Trove innovative?

The National Library has acknowledged that changes in user expectations, technology and the wider environment mean that some users do not expect to be passive receivers of information, but rather contributors and participants in information services, and thus able to share their ideas and information. Trove enables this by giving users the ability to tag, comment and correct content, and organise their results. People can upload their own images into the service. Users are encouraged to give feedback and engage with content and this is driving the future development of the service.

Where is the content coming from?

The content is coming from the National Library of Australia, over 1,000 libraries around Australia and from other cultural and educational organisations. International digital collections of relevance are also being harvested. By early 2010 the service had a body of content of millions of items (much of it digital), which will grow as the service develops and more organisations contribute data.

Content scope

  1. Trove is a complete replacement for the free Libraries Australia search service, whose goal was to provide free simultaneous access to the nation‟s catalogues, a streamlined capability for borrowing a book from any library, or purchasing from copying services and booksellers.
  2. Trove has built on this to provide new content such as digitised early Australian newspapers, and information about people and organisations.
  3. Trove supports collaboration in existing National Library services by including:
    • the National Bibliographic Database – MARC records for items
    • Picture Australia – Dublin Core records for digitised images
    • PANDORA – web sites archived since 1996
    • newspapers 1803-1954 – articles from major metropolitan dailies
    • Music Australia – MARC records for sheet music collections
    • the Register of Australian Archives and Manuscripts – finding aids
    • authorised names of people and organisations.

    The separate interfaces for the Register of Australian Archives and Manuscripts (RAAM), People Australia and the free Libraries Australia search service have already been integrated into Trove.

    The National Library discloses digitised content from its own special collections through Trove, and using persistent links, this content can be featured in „virtual‟ collections about iconic Australiana, events or people.

  4. the National Library will continue to identify and select content to augment other sources already added, such as:
    • full text content from online public domain books such as the Open Library and the Hathi Trust
    • reviews from Amazon
    • Australian Research Online, which contains links to scholarly research generated by university libraries, federal and state government departments, and ejournals hosted by a variety of Australian knowledge organisations
    • tables of contents and sample book chapters from the Library of Congress
    • rich biographical profiles from the Australian Dictionary of Biography Online, the Australian Music Centre, the Australian Women‟s Register, the Dictionary of Australian Artists Online, and the Collections Australia Network (which represents museums)
    • OAIster (a service hosted by OCLC) which contains more than 20 million records for scholarly materials sourced globally from universities and other research organisations.

    Details of content sources are kept up-to-date at trove.nla.gov.au/general/faq#where.

  5. The service mixes fulltext and images with records linking to borrowing and buying options as well as supporting a choice of finding an item in a particular format – a book, sheet music, sound samples, a photograph, a newspaper article, an old web site, a biography, an interview, and an eprint.
  6. Trove engages with all Australians through service features such as the ability to add comments and tags to material they have utilised as well as providing a workspace to come back to, a function to download citations, and the correction of digitised text. These features were trialled in the Australian Newspapers service.

    Other features include the ability to specify (after registration) a range of preferred libraries for borrowing from, and checking the copyright status of items. They were introduced in Libraries Australia.
  7. Trove will continue to be upgraded. Frequent advice about this is posted on the site at trove.nla.gov.au/general/completedlist. It is also possible to sign up to an announcement list, and in 2010, one or two unmediated social networking spaces which will be hosted by the National Library to allow both the general public and collaborators (data contributors) to discuss issues and future directions.
  8. The Trove interface has been developed and refined using feedback from the public. Trove provides a significant new way to expose your collections to the public beyond your current pathways. The ability to create lists of items found, to represent themes or just share items with others, provides a means to feature your collections in compact form.

Benefits of Trove to collecting organisations

Trove provides a significant new way to expose your collections to the public beyond your current pathways. Trove is the National Library‟s marketing focus for 2010. It is therefore likely to have a significant complementary impact on other organisations‟ discovery services and usage of collections if the options described below are implemented as outlined. The expected opportunities are as follows:

  • the Trove interface acknowledges existing services by including brands in each item display;
  • reference referrals received by a National Library help desk are forwarded to a relevant organisations (using a generic email address unless otherwise advised);
  • automated search referrals, as described below, which will increase traffic to linked web sites, providing a flow-on benefit of the participating institutions. Referral statistics collected at domain name level are publicly available at trove.nla.gov.au/system/stats#links;
  • the identification and inclusion of public domain sources of extensive fulltext mean that other collecting institutions don‟t have to acquire it, which saves processing costs (both manual and IT), and in the future, shelf space;
  • the lists function can be used as an alternative mechanism for providing subject guides;
  • the anticipated social networking spaces allowing the general public to discuss their expectations of library services;
  • the comments from the general public embedded in Trove about individual items are publicly available at http://trove.nla.gov.au/recentComments.

    The comments contain suggestions on where to buy books, for example
    Book: The Green Horse Owners Guide to Owning a Horse. Comment supplied: “this work is available from Folio Books Albert St Brisbane.”

    stories about the item, for example,
    Image: Hotel Francis, Shelly Beach, Caloundra, date unknown. Comment supplied: “And was demolished in December 2009.”

    and corrections to metadata or text (particularly for newspaper articles) and ratings, for example,
    Newspaper article: Altercation between bailiffs. Serious Complications, (News), The Brisbane Courier (Qld:1864-1933), Friday 19 February 1892 page 3. Comment supplied: “The last 5 lines could not be corrected.”

As part of the national information infrastructure, Trove provides a technical platform which can be utilised in ways to enrich the discovery services of organisations with significant collections. The National Library is ready to assist collecting organisations to undertake more intensive activities to increase awareness and use of their unique collections. Such opportunities include:

  • the ability of organisations to obtain a list of unique identifiers for individuals from Trove, which can then be used to fetch biographies back into local discovery services in real time, as well as augment local biographical services;
  • the ability of Trove to accept minimal records or modest file formats (such as EXCEL spreadsheets) through the Libraries Australia Record Import Service for special collections which are still to be discovered by the public;
  • support for libraries to disseminate advice and guidelines for the digitisation and inclusion of regional newspapers into a single location for discovery;
  • re-tailoring of local discovery services to exploit the Trove search engine;
  • the ability to identify gaps in collection development, particularly across different format types such as a manuscript and a published version of a work;
  • Trove marketing materials, and messages/presentations about Trove, are available for reuse by any organisation at any time from http://trove.nla.gov.au/general/marketing.

PART 2 OPTIONS FOR UTILISING TROVE

How to refer to Trove on your website

  • As outlined in Part 1, all references on public websites to the Libraries Australia free search service should be changed to Trove.
  • All professional references should list Trove and Libraries Australia – Trove for public spaces and Libraries Australia for workflow services available to staff responsible for cataloguing, interlibrary loan or document delivery.
  • On some occasions, „privileged‟ community members such as researchers or authors may have individual access to Libraries Australia (set up by Libraries Australia member organisations), so that they can use services such as interlibrary loan without an intermediary. The National Library recommends pointing out both services to privileged members.

Web sites linking to Trove

As part of the rollout of Trove, advice was given to all member libraries via the Libraries Australia web site, discussion lists and newsletter on changes to service delivery. For example, where a Libraries Australia search box has been embedded in a web site, it was automatically upgraded with a Trove search box, and people using the free Libraries Australia search service were automatically redirected to Trove. Many organisation web sites contain many pages which list links loosely grouped under a catch-all heading such as Catalogues or Search databases. Trove is not a catalogue, it is primarily a search engine. Descriptions of Trove are available at trove.nla.gov.au/general/marketing.

Organisations which have created entries in Wikipedia have sometimes included links to records in the National Bibliographic Database. While the National Library has been able to automatically redirect some to Trove, any hard-coded text can only be changed manually by the organisation which created it.

The first group of options discusses possibilities for what the public sees. These examples were captured in January 2010 and illustrate where small, swift changes should be made.

 Libraries Australia – Perform simple and complex searches across a wide range of databases, including the National Bibliographic Database.

Trove uses new features such as relevance ranking and faceted browsing to obviate the need for complex searches. We recommend updating the wording and linking to Trove instead.

Example of how to update Libraries Australia links to Trove

Libraries Australia is more than just a catalogue. It has connected people to collections held by all of Australia‟s significant libraries and other organisations which have welcomed outreach for their collections through this platform. They are now all integrated into Trove. We recommend updating the wording and linking to Trove instead.

Example of how to link to Trove

Note that while the word Australia appears as part of the brand for the new service, it is not part of the name of the service, and should not be included in links.

Example of the wording to use when linking to Libraries Australia

The link in this search result naming Libraries Australia actually searches Trove because it is redirected. We recommend updating this type of link in any OPAC software.

Example of text to be used to replace old Kinetica information

The Kinetica brand was replaced in 2005 by Libraries Australia. We recommend updating the wording and linking to Trove instead.

Implementing Trove search on your website

The second group of options describes mechanisms for the Trove platform to interface with other services (known as machine-to-machine interfaces). Once implemented, they usually do not require further attention unless an environmental condition changes, such as a change to a web domain name. Some have already been developed, such as a Google sitemap.

The National Library has facilitated the harvesting of records by Google from the Trove repositories using a new Google sitemap. If an organisation has provided records or content via the single services mentioned above, then this content will be made available to Google for harvesting. Note that while Trove gathers in updates quickly (hourly, daily), Google itself may take several months. The Library continues to research ways to streamline this and responds to new approaches implemented by Google and other search engines.

At individual organisation level

Machine-to-machine interfaces already available to individual organisations include:

  • search box
  • canned searches
  • deep linking
  • Search and Retrieve in a URL (SRU)
  • your browser.

Search box

Prior to launch, the National Library alerted all member libraries that the Libraries Australia search box would be automatically replaced with the Trove search box. However, any text associated with the search box provided by hosting sites may need to be modified. This site has implemented it correctly:

 Image of a website that has implemented the Trove search box correctly

 

The embedded search box is a set-and-forget way of encouraging searchers to use this Australian search engine, see http://trove.nla.gov.au/general/marketing#searchbox.

Canned searches

Aficionados of National Library innovation have already created little applets which exploit access to Trove from their own web sites. One is for canned searches. Trove URLs are structured for easy exploitation, see http://trove.nla.gov.au/general/searchURLs.

For example, http://trove.nla.gov.au/result?q=WORD+WORD

Or to restrict a canned search to a zone in Trove such as the Books, Journals, Pictures and Photos view http://trove/nla.gov.au/picture/result?q=victoria

Deep linking

Deep linking is a more refined canned search. It uses a unique key such as an ISBN, a record identifier or author and title details to query a local discovery service directly. Results sets are shorter, and should usually contain only one item or record, immediately presenting the searcher with local availability information such as conditions for borrowing. To set up deep linking in Trove (and Libraries Australia), details can be sent to the Libraries Australia team via a form at www.nla.gov.au/librariesaustralia/search/deep-linking.html.

Search and Retrieval in a URL (SRU)

This protocol has been developed as a web response to Z39.50, which was not architected to work in web environments. The SRU protocol has already been implemented to gather in biographical data and send out unique people identifiers to organisations contributing to the People Australia program, but requires a particular data schema to be implemented by the collecting site. Further information is available at trove.nla.gov.au/general/aboutPeople.

Your browser

Trove may be added to any browser as a default search target, trove.nla.gov.au/general/opensearch

 Image of setting Trove to your default search target

 

Future machine-to-machine interfaces to be scoped include:

  • an Application Programming Interface (API). The technical specifications for the API are to be the subject of further consultation. The API will provide access to the data stored in Trove, for harvesting and data mining purposes, in different schemas such as Dublin Core and Encoded Archival Context. Files of MARC21 records will remain available through Libraries Australia only. Exchanging annotation files is also being considered;
  • OAI-PMH (Open Archives Initiative Protocol for Metadata Harvesting) and SRU Record Update will be considered.

At the network level

All of the options described above are deployable at individual organisation level. Trove can also be used to realise new networking options. Three suggestions are made here but they should not be considered the only possible ways to exploit Trove.

  1. The Trove function to allocate tags could be used within small communities to share discovery, or link like items together for reuse;
  2. The Trove function to choose preferred libraries could be used in the future to create a virtual shared catalogue. This is true for any group of libraries which have an interest in affiliating with each other. The flexibility of the Trove interface design would allow some targeted branding of such a „catalogue‟.

    To date, this function works at the individual level. The function will have greater applicability within specific sectors at a network level when combined with the proposed API development. A departmental library could create a set of libraries of interest. Such flexibility could also be of value for a consortium such as the Queensland Government Libraries Consortium, to create a shared virtual discovery experience.

    Image of information about your interface in Trove

    A cultural “portal” could be created in the same way. However, it can only be built if all of the organisations of interest are members of Libraries Australia.

    Image of Your libraries interface in Trove

  3. The Trove function to create lists will be used to replace subject guides in Australia Dancing, trails in Picture Australia and themes in Music Australia. Lists have the additional advantages of allowing tailored descriptions of any length, do not require Trove team assistance to create, and can be modified at any time. The function to co-author will be added. This example is at trove.nla.gov.au/list?id=761.

    Example of lists in Trove

Conclusion

Trove has only recently been launched, so as the service builds momentum new opportunities for its exploitation will become available. In summary, some options for engaging with Trove in 2010/11 are:

  • using Trove features such as lists and annotations to educate communities in web 2.0 practices (described here http://virtuallyalibrarian.com/2010/06/09/troves-new-list-feature-a-great-tool-for-educators/);
  • encouraging wide use of the service, to generate flows back to local services;
  • continuing to contribute data through existing program pathways such as Libraries Australia, Picture Australia, Music Australia, People Australia and Australian Newspapers;
  • improving data contribution by maintaining the currency and accuracy of holdings information and URLs;
  • generating interest in the API to support data mining;
  • establishing shared library profiles for collection management purposes.

The National Library is happy to provide further advice on these options, and welcomes any suggestions for new public engagement and further exploitation of the national information infrastructure offered by Trove at any time.

TROVE MANAGEMENT

Collaborative Services Branch at the National Library manages Trove and the underlying programs that now give content to the Trove repositories. These are: Libraries Australia Customer Services (with the assistance of the Database Services team), the Picture Australia program, the Music Australia program, the People Australia program and its associated ANDS-funded Party Infrastructure Project, Australian Research Online, the Australian Libraries Gateway, the InterLibrary Resource Sharing Directory, and Australian Newspapers.

Trove enquiries are received by the Libraries Australia Help Desk for appropriate dissemination or response.

REFERENCES

  1. Developing Trove: The Policy and Technical Challenges. February 2010, Cathro, W., Collier, S.
  2. Resource Sharing and Innovation Division Strategic Plan. July 2010, Cathro, W.
  3. Libraries Australia and Trove – the public speaks” in Libraries Australia newsletter Issue 6, Campbell, D. 30 April 2010