National Library of Australia
Trove Party Infrastructure
Revised 10 Feb 2012
This work is licensed under a Creative Commons Attribution 3.0 Australia License
The party infrastructure was developed by the National Library of Australia as the People Australia Infrastructure to support the persistent identification of people or organisations.
This development was supported by the Party Infrastructure Project (Jan 2010 to July 2012) funded by the Australian National Data Service (ANDS). The core aim of the Project was to adapt the National Library’s People Australia infrastructure so that the research sector could contribute and manage their party records and obtain NLA-issued persistent identifiers for their records about researchers and research organisations. The Project o bjective was to develop a sustainable infrastructure for publicly identifying parties (people and organisations) in the research sector.
An example of the problem the party infrastructure was attempting to solve is demonstrated by the following examples of public information about Tim Flannery:
Flannery, Tim F. (Tim Fridtjof) (1956-)
He is also known as:
Flannery, Timothy F.
Flannery, Timothy Fridjof
And he may also be known as:
He may have multiple author identifiers as a result of his publications:
e.g. Scopus Id, Researcher Id, Libraries Australia Name Authority File Id, VIAF Id, etc.
The solution as provided through the party infrastructure is:
- One identifier, issued by the NLA
- A public identifier.
- A persistent identifier.
- Managed in the public domain.
- Using public information.
- And is available in the NLA’s free public interface Trove http://trove.nla.gov.au
Through the National Library’s Trove Party Infrastructure, different organisations can contribute a record about the same person or organisation and these records are grouped under the one identity record which has one NLA persistent identifier.
In this way, a public profile of the person or organisation is built up and has the potential to include all of their public identifiers and publications. Records from each contributor are kept whole and grouped, or co-located, under the one persistent identifier.
An example of the record structure is:
Which is expressed in the Trove People and Organisations zone as:
The basic workflow for the party infrastructure is as follows:
- The contributor has party records in either EAC-CPF or RIF-CS metadata format and can make these records available for harvesting by the NLA via OAI-PMH.
- The contributor contacts the NLA via the Trove ‘Contact us’ form on the Trove homepage ( http://trove.nla.gov.au ) and requests that their party records be included in the Trove Party Infrastructure.
The contributor advises the NLA:
The Contributor name that is to be displayed in Trove
- The Contributor’s ISIL
- The number of records they wish to contribute
- The format of the records (RIF- CS 1.2 or EAC-CPF)
- The base URL (and set/ group if necessary) for OAI harvesting of their records
- The NLA checks and advises the contributor if they need to sign a Trove Memorandum of Understanding before harvesting the contributor’s records into a test environment of the party infrastructure.
- The Contributor is added to the List of Trove Contributors waiting for processing and a priority is assigned according to the NLA’s available resources and workload. The time frame for the work will be discussed with the Contributor.
- If the Contributor does not have an ISIL, they should organise to get one. Details about ISILs and how to obtain one can be found in another section of this this Guide.
- When the work is ready to start NLA will arrange to get the Contributor’s name and ISIL added to Trove and the Trove Identities Manager (TIM) for the test and production systems.
- When the Contributor has been set up in the NLA’s Party Infrastructure, the harvesting process of the records via OAI-PMH will commence.
- In the first instance a test harvest will be done in a test environment to check all systems are working properly and the records can be obtained via OAI, and record format is checked. If there are issues with the harvest of data at this stage NLA will liaise with the Contributor to resolve any problems.
- When records are harvested, they go to the identity service where automatic matching takes place. In the identity service, matching rules are applied to harvested records that compare the harvested record with identity records that already exist in the infrastructure. A record that is automatically matched is co-located with the matching identity record and adopts the NLA identifier. A record that fails automatic matching is made available for human review via the Trove Identities Manager.
- Records are then harvested to the TIM Beta and the Trove Test system. The Contributor will be advised of the results of the successful harvest and load and that their records are able to be viewed in the Trove Test system.
- If the Contributor will be managing their unmatched party records in TIM, they need to sign up for a Trove user account and advise the NLA of their Trove registration details. Information on the Trove signup process is available in another section of this Guide
- NLA will then arrange for them to be given access to TIM (Beta and Production) for this Trove registration so they can do a human review and match of their unmatched records. Note that records that fail automatic matching will not appear in the party infrastructure until they are reviewed and hand matched in the Trove Identities Manager.
- The NLA will harvest the records into the Trove and TIM production environment where NLA persistent identifiers are issued.
- A schedule for the regular harvesting of the records is agreed with the Contributor and this is set up in the NLA’s Harvester.
- The Contributor will have access to TIM where they can manage their unmatched records, to match to an existing identity or to create a new identity.
- The Contributor can choose to get their party records, and assigned nla party identifiers, back from the party infrastructure via the NLA’s APIs - SRU, OAI or by searching Trove.
The main components of the party infrastructure are:
The NLA Harvester
Harvests records from contributors to the party infrastructure via the OAI-PMH.
The identity service
Applies matching rules to harvested records that compare the harvested record with identity records that already exist in the infrastructure. A record that is automatically matched is co-located with the matching identity record. A record that fails automatic matching is made available for human review via the Trove Identities Manager.
Trove Identities Manager
Formally known as the Party Administration Tool or the People Australia Data Administration Module, this is an online Tool where NLA staff and external contributors can view the records which failed automatic matching through the identity service and match or create identity records by hand.
Trove People and Organisations zone
An online interface in which party records in the infrastructure can be viewed by humans.
Enables party records is the infrastructure to be obtained using SRU standard XML search protocol.
Enables party records is the infrastructure to be obtained using the Open Archives Initiative – Protocol for Metadata Harvesting (OAI-PMH).
Records in the party infrastructure are stored in EAC-CPF format. A conversion to this format from other record formats such as RIF-CS are made via an XSL stylesheet in the NLA harvester.
Before becoming a contributor to the party infrastructure, the following preconditions need to be met:
- The contributor has party records to contribute.
A ‘party’ refers to a person or an organisation. A party record describes a person or organisation and may contain some or all of the following information: identifiers, name, name variants, birth and death dates, occupation, field of research code, subjects, biographical information, associated people or organisations, publications or websites by or about the person or organisation. Please see the section in the ‘Guidelines for party record content and format’ in this Guide.
- The contributor has party records in RIF-CS or EAC-CPF format.
The Registry Interchange Format - Collections and Services (RIF-CS) Schema was developed as a data interchange format for supporting the submission of collections metadata to the ANDS Collection Registry. The party infrastructure can accept records in RIF-CS 1.2.0. More information and schema documentation can be found on the ANDS website at http://www.ands.org.au/resource/rif-cs.html
The NLA will advise when support RIF-CS 1.3 will be provided.
The Encoded Archival Context – Corporate Bodies, Persons and Families (EAC-CPF) is the record format used internally, and for record exchange, in the party infrastructure. The EAC-CPF Schema is a standard for encoding contextual information about persons, corporate bodies, and families related to archival materials using Extensible Markup Language (XML). The standard is maintained by the Society of American Archivists in partnership with the Berlin State Library. More information and schema documentation can be found at http://eac.staatsbibliothek-berlin.de/ eac-cpf-schema.html
- The contributor has made their party records available for harvesting via OAI-PMH.
Party records are harvested from contributor’s websites using the Open Archives Initiative – Protocol for Metadata Harvesting. OAI-PMH is a web-based protocol established in 2001 to define a standard way to move metadata from point A to point B via the Internet. The OAI provides rules and a framework for sharing descriptive metadata, both for making metadata available and for acquiring metadata records once they are made available. The NLA uses a custom-built harvester application to issue OAI requests and collect metadata from various repositories and other collections of data. For more information on OAI-PMH see http://www.openarchives.org/pmh/ .
- The contributor has signed a Trove Memorandum of Understanding (MoU) where required.
Some contributors to the party infrastructure may be required to sign a Trove Memorandum of Understanding (MoU) before their party records can be harvested and made available in Trove.
- The contributor has an ISIL (International Standard Identifier for Libraries and Related Organisations).
An ISIL is an International Standard Identifier for Libraries and Related Organisations (see http://biblstandard.dk/isil/ ). It is a unique number that can be assigned to every library and related organisation in the world. An ISIL is alphanumeric and takes the form of a country code prefix, then a hyphen, and then an identifier issued by that countries national library authority.
The National Library of Australia has the authority to assign ISILs to libraries and related organisations in Australia. These take the format of country code ‘AU’ for ‘Australia’ then a hyphen (-) and then an identifier. The identifier issued by the National Library is a National Union Catalogue (NUC) symbol. NUC symbols are widely used by the Australian library and research community for identification and communication within the national and regional networks. For example, the ISIL for Monash University is AU-VMOU where AU is the country code and VMOU is the NUC symbol. The ISIL for the Australian Women’s Register is AU-VU:AWR.
To check or obtain an ISIL, a contributor needs to:
- Go to the Australian Libraries Gateway website http://www.nla.gov.au/libraries/ and use the search options to check if their organisation has a NUC symbol.
- If their organisation has a NUC symbol, simply add the prefix AU then a hyphen (-) to formulate the ISIL. For example, the NUC symbol for Griffith University is QGU. The ISIL for Griffith University is ‘AU-QGU’.
- If they can’t find an entry for your organisation in Australian Libraries Gateway, or they are unsure of their NUC symbol, they should contact the Trove team.
- The contributor has access to, and is familiar with, the Trove Identities Manager.
Before using the Trove Identities Manager to create and match records by hand, the contributor should understand the purpose and function of the Trove Identities Manager. They should consult the current version of the Trove Identities Manager Guide .
The Trove Identities Manager (TIM) allows for the maintenance of records in the party infrastructure. It enables authorised staff from institutions and organisations who contribute party records to the infrastructure, and staff from the National Library, to manage these records.
The Trove Identities Manager:
- is web-based and available via the internet;
- was built in-house by the NLA;
- is separate but linked to the NLA Harvester, the application used to get party records into the party infrastructure.
- enables authorised staff in research sector, organisation and at the NLA to manage records and identities about people and organisations for the party infrastructure.
A Trove user account login is necessary for a Contributor to access their unmatched records in TIM.
A Contributor can review and manage their unmatched records, either to match an existing identity or to create new identity.
- Each user account (login) is associated with one or more contributor collections in TIM.
- A user of TIM will only be able to see and action unmatched records that are associated with the collection(s) assigned to their user account.
- All users of TIM can see all identity records regardless of the collection(s) assigned to their user account but they can only action records assigned to their user account.
Trove Sign up and user account profile
A contributor needs to sign up for a Trove User account via the Trove home page. When the Contributor has their User Account details they should notify the NLA of the Contributors name, the ISIL, their User name an email address, so NLA can arrange for them to have access to their unmatched records in TIM.
A Trove User login is required to access the Trove Identities Manager. To To sign up to Trove:
- Go to the Trove home page http://trove.nla.gov.au/ and click on the “Sign up” link
- From this page you can Sign up as a registered user of the NLA’s Trove service
- An email will be sent to your nominate email address and this email will contain the details an link to activate your Trove User Account. You should active both your user accounts in the Trove production system http://trove.nla.gov.au/ and the Trove Test system http://trove-test.nla.gov.au/
- The same Contributor’s Trove User Login is used for both the Trove Test and Production systems, and for the Tim Beta and Production systems, when access has been set up.
Note: If you do not active the Trove User account it will lapse after 3 days, and the account will not be established.
Note: NLA does not store your password. If you forget your password you will need to follow the instruction in the Trove “login” page to reset the password.
A party record should include as much detail about the person or organisation described in the party record as possible. This information may include:
- name variants
- date of birth/date of death
- field of research
- occupation or function
- related people or organisations
- publications by the person/organisation
- publications about the person/organisation
When describing a party using a metadata schema such EAC-CPF or RIF-CS, it is recommended to:
- Split the name into parts and describe those parts.
- For party records which describe a corporate body or group, order of parts is particularly important and should start with the superior corporate body or group and end with the subordinate.
- Include as much information about the party as possible.
Including as much descriptive information about the party as possible, and following the guidelines for describing the information, will assist in improving the rate of automatic matching and creating a richer party infrastructure for users.
Records can be easily obtained from the party infrastructure by searching for the records in Trove or by using one of the party infrastructure APIs:
The OAI Repository is available at http://www.nla.gov.au/apps/peopleaustralia-oai/ . Record formats available are EAC, RIF-CS and Dublin core.
2. SRU (Search and Retrieve via URL)
SRU provides a standard web-based text-searching interface, utilises a rich query language called CQL (Contextual Query Language) and provides records in XML format. The SRU interface is available at http://www.nla.gov.au/apps/srw/search/peopleaustralia . Record formats available are EAC, Dublin core, RIF-CS and ATOM.
Detailed documentation of the NLA Party Infrastructure SRU Server is available in PDF Format .
The OpenSearch interface description file is available at: http://www.nla.gov.au/apps/srw/peopleaustralia-os-description.xml Record formats available are RSS and ATOM format. OpenSearch supports basic keyword searching and provides records in XML format.
- Trove search and web feeds
Party records for a Contributor can be seen in Trove by searching in the Trove People and Organisation Zone for the Contributors ISIL, e.g. use the search “AU-VANDS” to get a results set of all the records contributed by the Australian National Data Service.
A Contributor can subscribe to the Trove web feeds to get notification to changes to the results of this search.
RIF-CS schema: http://www.ands.org.au/guides/content-providers-guide.html
RIF-CS examples: http://www.ands.org.au/guides/cpguide/cpgparty.html
EAC-CPF schema: http://eac.staatsbibliothek-berlin.de/
EAC-CPF examples: http://www3.iath.virginia.edu/eac/cpf/examples/list.html
- NLA Party Infrastructure Project wiki: https://wiki.nla.gov.au/display/ARDCPIP/
- Trove Home page http://trove.nla.gova.au
Use the “Contact us” form on this page to contact the Trove Support Team.
- Trove People and Organisation Zone About page http://trove.nla.gov.au/people/