Technical ecosystem

Trove system overview

Trove helps patrons find, access and enhance the collections of Australia’s Libraries, Museums, Galleries, Archives, Research institutions and other collecting organisations. It does that by bringing together their collections into a single place, so patrons don’t need to search different websites and can find out about organisations they didn’t even know existed.

Trove isn’t one single piece of software, it’s an ecosystem where each piece plays an important role in delivering the service you know and love. This includes:

  • A discovery service, where patrons can freely search for, find and access all of Trove as well as make contributions including correcting text, adding tags/notes and creating Lists all within their internet browser.
  • Digital repositories providing a home for the full text items within Trove. These repositories help Librarians and Archivists care for the needs of digitised and born-digital material throughout the digital lifecycle, from first collecting/digitising the item right through to delivering it to patrons in Trove.
  • An aggregator to store metadata from partner organisations, helping patrons find items that live outside Trove and view the items online. This links Trove to hundreds of other digital repositories at other organisations.
  • A co-operative cataloguing database where Librarians share the effort of describing their collections and letting patrons know what material they hold on their shelves.
  • A directory of collecting institutions that connect patrons directly with staff at partner organisations by sharing contact details including address and opening hours.
  • A supporting search service, which is tasked with indexing the digital repositories and the aggregated metadata then providing search and browse capabilities to patrons through the discovery service.
  • An Application Programming Interface (API), that makes the discovery service available for re-use by machines.

See the System Overview for an illustration of Trove’s ecosystem of interlinked applications, data stores and data flows.

For more of the technical details read on.

Trove Discovery Service

The Trove discovery interface is entirely browser-based and was re-built from scratch in 2020. It’s based on Vue.js technology. Vue.js is an open-source Javascript framework for building user interfaces and single page applications. It features an incrementally adaptable architecture that focuses on rendering and component composition, along with routing, state management and build tooling. Supported by officially maintained libraries, vue.js makes Trove easier to develop in the future.

Digital Repositories

Trove has three digital repositories that house different types of content. Each repository provides special tools to help Librarians and Archivists care for the material, and for patrons to interact with items. All three systems have features for collection/digitisation/online deposit, description, quality assurance review, storage, preservation, rights management and delivery to patrons.

Trove Digital Library is the repository for National eDeposit (NED) born-digital items as well as collaboratively digitised journals and books. In addition, it holds the National Library of Australia’s digitised collections such as photographs, maps, oral histories and diaries. Trove Digital Library works closely with other software for some of its standardised functions. For example to manage archival collections it integrates with ArchivesSpace and to undertake digital preservation in integrates with Preservica. Items from the year 1530 onwards have a home in the Trove Digital Library.

Newspapers Content Management is the repository for digitised Australian newspapers. It’s a built-for-purpose system to manage all the backend processes associated with mass digitisation of newspapers from microfilm, as well as the online delivery of those articles to patrons. It contains metadata, images and text for millions of pages of Australian newspapers published from 1803 onwards.

The Australian Web Archive is the repository of Australia’s freely available websites. Librarians are able to capture and store webpages on a yearly basis and more frequently, preserving them for delivery to the public now and in years to come. It holds snapshots of websites in the .au domain, from 1996 to today and is continuing to capture the internet.

All 3 repositories are accessed through the Trove interface. They have been built by the National Library of Australia in Java, integrating with open source and commercial off-the-shelf products where available.

Digitise-on-Demand

Copies Direct allows patrons to find items in Trove and purchase a copy. When a patron does that, they’re also sponsoring the item to be digitised and made available online (rights permitting). This gives all other patrons free access now and in the future. After a patron enters a request for an item a member of the National Library of Australia’s staff finds the item on a shelf and makes a digital copy. This is then made available on Trove for all to see and stored in the Trove Digital Library repository. A copy is also sent to the original requester.

This service was partially built in-house using Java, and works with the Relais software to make sure all requests are handled, whether they’re from patrons or partner libraries.

Aggregator

The NLA Harvester regularly visits the websites and repositories of partner institutions and makes copies of their metadata to bring back into Trove. That metadata is then made available in Trove’s discovery service, allowing patrons to find items from hundreds of different collections in a single set of search results. When a patron finds an interesting item they can follow the link off to the institution’s website and view the item from their repository.

Like most of our systems, the NLA Harvester was built in-house on Java technologies and first came online in 2008. The Trove Identities Manager relies on SOLR and Java, and was launched in 2011.

You can read more about the technologies the NLA Harvester can interoperate with on the Technical Specifications in our Partner section.

The Australian National Bibliographic Database (ANBD)

The ANBD is where Australian Librarians come together to co-operatively catalogue their collections. Partner libraries contribute their bibliographic records for the published and unpublished material in their collections – items like books, DVDs, newspapers and so on. Librarians around the country can then download the detailed descriptive record and re-use it in their own catalogue without having to create it themselves, saving everyone time and money.

Librarians also add “holdings” to the database. This lets patrons and other Librarians know that they hold a copy of that item in their collection, and it might or might not be available for viewing or borrowing. This helps patrons find an item in Trove and then approach a library to view it, and allows libraries to carry out Inter Library Loans and Document Delivery.

The ANBD was originally launched in 1981. Today it is a series of Java applications powered by a combination of open source SOLR and MySQL, as well as the proprietary CBS and VDX applications.

Directory of Collecting Institutions

The Australian Libraries Gateway is a directory of thousands of collecting institutions and their contact details. While originally envisioned to hold the content details of all Australian Libraries, it has grown to also include Trove partners who are Galleries, Archives, Museums, Research Institutions and community groups.

Collection owners regularly update their details, and the information is used to power display features in the Trove discovery service including collection names and availability to the public. Partners can also visit the directory directly.

First launched in 1998 the directory was built in-house and is a combination of Java and a MySQL database.

Trove Search service

The Trove Search service is the central meeting place for the digital repositories, aggregator and co-operative cataloguing database. All the descriptive data is copied from those separate homes into the one search service where it’s made ready for patrons to type in their keyword searches.

The search service is like a farmer’s market – all the different produce is brought together in one place for shoppers to choose from separate stalls selling fruit and vegetables, or plants, or bread. In the same way Trove’s search service brings together all the separate data types, with aggregated data in one place and digital repositories in another. Different features are

provided depending on the type of material, though some features like keyword search are common across all parts of the service.

Launched in 2010 the Trove Search service was built in-house and is based on SOLR.

Application Programming Interface (API)

The API is an exit out of the Trove ecosystem. It allows any member of the public to pull data out of the Trove Search service and re-use it in their own projects, whether that’s a website, detailed data analysis and research or even their own digital repository.

Much of the data in Trove is available through the API. Where the Trove discovery service is intended for patrons to view, the API is intended to be read by machines and computer programs. You’ll need an API key to get started, but don’t worry, the process is automatic for non-commercial projects.

The API was first made available in 2011 and was significantly upgraded in 2019. It was built in-house with Java. Find out more through Using our API.