Best practice for creating party records

Structure and meaning of 'party'  |  Best practice for creating a party record  |  Example RIF-CS  |  Contributing party records to Trove

Creating a Party record

The purpose of a party record in Research Data Australia is to support discovery of research data collections, and to provide context to those collections.

ARDC Party Infrastructure Project

ANDS is collaborating with the National Library Australia to provide infrastructure for describing parties using the Trove service. All ANDS partners can use this infrastructure to create party records which can be harvested into Research Data Australia from Trove. More information

Step 1: Should I create a party record at all?

At least one party record needs to be related to each collection described in Research Data Australia (see Metadata Content Requirements).

You do not need to create a party record if:

• A suitable party record already exists in Research Data Australia.
  
• A suitable party record already exists in the National Library of Australia's Trove.

If a record is already available for your party, you can link to it by including either the Research Data Australia party record key or the NLA Trove party identifier in the related object element of your records. You do not need to create a new party record.

• You want to describe a party as the subject of a collection. Use the subject element in the collection to record the party as a subject, ideally including a term identifier from a name authority.
  

You need to create a party record if:

• There is no suitable party record already in Research Data Australia or Trove. How to find out
  

• A party record already exists in Research Data Australia, but the existing record doesn't say enough about your party and you wish to add value to what is already there.
  You might need a more comprehensive description, a more up-to-date description, or a description better reflecting your institution's concerns.
  You will need to provide your own description of the person or group in a separate party record. You cannot edit other partners' records, including the National Library of Australia’s (NLA) records. See Step 10 Identifier for how to make sure that both your record and the existing record are treated as describing the same party.

Important: Most partners will need to contribute party records, because many researchers may not have records yet, or the records that do exist may be inadequate.

You might need to create a party record if:

• Inclusion would not contribute to discovery or provide context for collections, but would support access to a collection by providing contact points.
  
• A party is only being included to provide context for other parties. Research Data Australia is not a party registry, so a party record should not normally be included in this situation.
  

Which party records am I responsible for creating?

• Create a party (group) record for your own research institution/centre.
• Create a party (person) record for a researcher from your own research institution/centre.
• Do not create party records for researchers external to your research institution. It is the responsibility of each research institution in collaborative projects to create party records for their own researchers. This approach has been adopted by ANDS because each research institution is the best authority on their own researchers, and can provide the most accurate, comprehensive and current information.
• International researchers can be included in the National Library of Australia (NLA) Party Infrastructure if contributors wish to do so. The NLA has no objection to universities contributing party records for their international researchers to Trove. Please note that the ANDS position is neutral, in that ANDS will neither recommend nor discourage this action.

ANDS suggests that you work with collaborating institutions to ensure that all researchers responsible for the generation of collections are appropriately acknowledged by having party records in Research Data Australia. The ARDC Party Infrastructure Project is intended to help partners access party records for external researchers.

How many party records do I need to create and link to my collection description?

At least one party should be related to a collection (see Minimum Metadata Content Requirements). If multiple parties can be related to a collection, the description should aim to link the collection to any party that will improve discovery substantively, and for which records can be maintained practically.

That means all active known collaborators on the project

• whose identities can be presented as well-managed Research Data Australia records, and
• who made a substantial contribution to the collection (so that it makes sense to discover the collection via the collaborator, and
• who can reasonably be given credit for bringing about the collection.

As a default, that would include all named researchers on the research grant application, but not support staff or research assistants (but not from other institutions, see Which party records am I responsible for creating?

What level of granularity should I create party records for?

Organisations are hierarchically organised, and you may have a choice of hierarchical level at which to represent your party (group). For example, you may link your collection to your research lab, to your department, your faculty, or your research institution. Which level of group is represented as responsible for collections is a matter of institutional policy.

Remember that Research Data Australia is not intended to represent group hierarchies. The default approach should be to represent the lowest-level group, with the most direct engagement with the collection. That means that the research lab or individual researchers are the parties of interest that need to be described and linked to a collection, in preference to departments or faculties. (The name of a research lab, for example, may include its superior body's name as part of its name, e.g. Budawang University, Frontiers of Chemistry Research Laboratory).

Research Data Australia groups an institution's information for display using the group attribute, rather than requiring connection of all collection records to an institutional party record. However, contributors may make such connections if they wish; institutional records for ANDS partners have been created to facilitate this.

Step 2: Is my party already described?

Search to see if your party is already described  before adding a new party record.

Where to search:

• Research Data Australia

  • Use the Advanced Search and limit your search to 'Parties'. The keys will be those created by the partner. Identifiers may vary.

• ANDS Online Services

  • You can search within data provider groups from https://services.ands.org.au/home/orca/search.php and limit your search to 'Parties'. You can only search published records or your own records of any status.
  • The following sets of records include authoritative party records.
    • "Australian Research Institutions"— party (group) records for institutions, created by ANDS. They have ANDS keys and include the NLA party identifier in the identifier element.
    • "National Library of Australia"— party (person) records created by NLA. They have NLA keys and identifiers may vary.
    • “Trove - People and Organisations”—party records created by NLA. The majority of records are party (group) records. They have NLA keys and the key is repeated in the identifier element.

• Trove

  • Search for matching records in Trove’s People and Organisations view. If you locate your party, identify the NLA party identifier for inclusion in your records (see Identifiers and Related Object information below for how to do this).
    
  • Not all parties in Trove will have an NLA party identifier. There are authors in Trove that are not represented with NLA party identifiers, but just as name queries, such as http://trove.nla.gov.au/people/Hall%2C%20David?c=book&q=creator%3A%22Hall%2C+David%22 ; do NOT include such query links as NLA People identifiers.
    
  • If there is no party identifier in Trove, this means you have to create your own party record.

Step 3: Provide values for the various elements of the party record

For common elements see the Content Providers Guide. The steps below discuss issues that are specific to parties.

Step 4: Type

Type is required. There are three party types, person, group and administrativePosition. Definitions of party types

Using the administrativePosition type

Administrative position is a kind of party where the position name and contact information are present but the identity of the party filling the role is not specified. 

Many data sources provide ANDS with party names such as "Data custodian", "Data Officer", or "Data Manager". There are over 1000 collection records (about 10% of the records in Research Data Australia) that contain role names for party records. Such role names are common in large data management environments.

• Role names need to be separated from person and organisation information provided to the National Library of Australia's Trove database, as role names should not be allocated party identifiers.
• Role names also need to be removed from displayed browse lists of researchers and research organisation names in Research Data Australia, as they do not assist discovery.
• Role names are useful when accessing collections but are simply noise when navigating during the discovery phase.

Best practice:

• Use type="administrativePosition" when describing a generic administrative role such as 'Data Manager' or 'Data Custodian' where no other party information is to be provided.
• Do not use for describing the relationships between named parties (people or organisations) and collections, activities or services.
• Do not use for describing a role filled by a named party (person or organisation); there should not be any related object links from a party of type 'administrative position' to any other party.

Disambiguation: A party of type "group" is not the same as and has no relationship with the "group" to which a Registry Object is linked for purposes of display in Research Data Australia. See Group for more information.

Step 5: Key

Keys are required. Keys must be unique in Research Data Australia. More information on creating keys for parties

Collection records connect to party records using the Research Data Australia party record key (see example below) or alternatively by including an NLA party identifier in their identifier element (see Step 10 Identifier).

Example:

        <relatedObject>
            <key>102.100.100/6676</key>
            <relation type="hasCollector"></relation>
        </relatedObject>

Step 6: Name

Names must be provided in name parts. Name part types are available for person names (title, family, given, suffix) and for group names (superior, subordinate). More information on party names

Only use date ranges if the name has changed over time and older versions of the name have been recorded in the metadata being provided, such as when a research centre has changed its name since the related dataset was created.

Step 7: Location

Party locations are included to enable users to contact the person or organisation. Relevant locations may include an electronic address or a physical address.

• An appropriate electronic address for a party (person) is the URL of the researcher’s web page(s), and/or an email address. A physical address for the researcher’s office/research centre might also be appropriate.
• An appropriate electronic address for a party (group) is the URL of the research centre’s web page(s). A physical address may also be provided.
  

Constraints

• Only use “Date From” or “Date to” for party location information if you need to describe a period of time during which the location information was current.
  Do not use spatial location in party records if you have a better form of location. Geographic coordinates are not as useful as electronic or physical addresses for contacting people.
  

Step 8: Coverage

Do not use spatial or temporal coverage in party records. Spatial and temporal coverage describe the locations in space and time to which research collections, datasets and activities relate: they are the location or time that something is about, not the location or time where something is. Coverage is not relevant to parties.

Step 9: Description

Use authoritative sources for party descriptions.

• A party (person) description is typically taken from an institutional profile web page for a researcher.
• A party (group) description is typically taken from an institutional web page.
  

The description of a party should provide users with context for collections. Information about the collecting researcher's experience and reputation is useful for evaluating collections.

Step 10: Identifier

If there is a public identifier for a party (person), ANDS strongly recommends providing that identifier as a party identifier. This may be a local username or staff ID, NLA Party Identifier (Persistent identifier), Scopus Author Identifier, or ThomsonReuters ResearcherID.

Include all relevant identifiers, as this helps in the disambiguation of similar names.

Where do I get identifiers from?

Search here:

NLA party ID  |  Scopus Author ID (from author search tab)   |  ThomsonReuters ResearcherID

Examples (fictional):

NLA Party Identifier example:

        <identifier type="AU-ANL:PEAU ">http://nla.gov.au/nla.party-482768</identifier>

In Trove this is displayed as: Persistent Identifier http://nla.gov.au/nla.party-482768

Scopus Author Identifier example:

<identifier type="uri"> http://www.scopus.com/authid/detail.url?authorId=7201847888
</identifier>

ThomsonReuters ResearcherID example:

        <identifier type="uri"> http://www.researcherid.com/rid/B-8721-2000 </identifier>

Local and URI examples:

        <identifier type="local"> researcher:3950184 </identifier>
        <identifier type="uri">http://www.myuni.edu.au/staffprofiles/s3799332</identifier>
        <identifier type="uri">http://www.myuni.edu.au/faculty</identifier>

Display of multiple party records in Research Data Australia

ANDS will treat multiple records as describing the same party, as long as they have the same authoritative NLA party identifier in their identifier field.

ANDS will not merge records describing the same party into the one record. Instead, in future development, records from different sources will appear together in the same display in Research Data Australia. The NLA record for the party will appear first, as the authoritative party record. As at October 2011, such records are displayed as ANDS Suggested Links. See example

 Step 11: Subjects

Provide a subject to allow Research Data Australia to associate a party with a research field, and indirectly with other collections in the same field.  Subjects are also used by Trove to help disambiguate researchers who have the same name. Subjects are not required, but are recommended in party records.

If subjects are added for party records, there should be at least one subject from the ANZSRC classification codes,  Fields of Research (FOR), Type of Activity (TOA), or Socio-economic Objective (SEO).

Reference: Australian Bureau of Statistics 1297.0 - Australian and New Zealand Standard Research Classification (ANZSRC), 2008.

Parties as subjects: if a party is the subject of a collection, include that information in the subject element for the collection, rather than linking the collection to a party record. For more information see Term identifiers.

Step 12: Related objects

Primary relationships

Use the primary relationship opt-in function in the the Data Source Account configuration page to link all records within your data source to a party record for your organisation (available from December 2011). More information about primary relationships

Bidirectional links

ANDS infers and displays bi-directional links between related objects in Research Data Australia. If a collection links to a party within the same data source, the party record does not need to link back to the collection; ANDS will display the inferred reverse link in Research Data Australia. If the party and collection are from different data sources, ANDS will only display the inferred reverse link if the receiving partner has opted in to allow bi-directional links. More information about bi-directional links

For manually supplied records, ANDS requires partners to provide links in both directions, to familiarise themselves with the link structures involved.

Related Collections

Parties must be linked to a collection, through "hasAssociationWith", "isManagerOf", "isCollectorOf", "isOwnerOf" or "enriches" relations. The most important relations of these are "isCollectorOf" (the party that takes credit for the collection), and "isManagerOf" (the party that is curating the collection, and can be contacted for further information).

Related Activities

If an activity record exists, it should link to a party record. This is to allow networks of researchers and collections to be clustered around research activities (through the "isParticipantIn" relation). This also makes it possible for users to get in touch with at least one party involved in the activity, as a contact point (through the "isManagerOf " relation).

Relations between parties and activities are only relevant if they improve collections discovery. For that reason, the other possible party – activity relations, "isFunderOf", "isOwnerOf" and "hasAssociationWith", should only be included if they improve the discoverability of collections. If you wish to include details of the funding relationship between a party and an activity in addition to creating the relation, include this information in the activity's description (note) element.

Related Parties

Research Data Australia is a collections registry; accordingly, relations between parties are not required or expected.

If party to party relations are considered important (such as "isPartOf", "hasPart" (group only) or "isManagedBy, IsManagerOf"), they are only relevant in Research Data Australia if they improve collections discovery.

Certain relations apply to person only, or group only. For example, "isMember of" applies to party (person) records only and "hasMember" applies to party (group) records only. Parties can fund a related party ("isFunderOf") or can be funded by ("isFundedBy") a related party.

Related Services

A party can own ("isOwnerOf") or manage ("isManagerOf") a service.

Step 13: Related Info

Include links to related information which is external to Research Data Australia.

An important source of related information is  an institutional web page which lists a researcher’s publications. This may be an on-the-fly web page generated from a research repository. This type of information is distinct from a researcher’s profile web page, given as an electronic location.

The references in the research bibliography of a researcher can also be listed individually under Related Info; but linking to the researcher profile is simpler.

Step 14: Existence Dates

Existence dates record the birth and dates for persons, or dates of establishment, incorporation, dissolution or other start and end dates for organisations. Existence dates support the ARDC Party Infrastructure. Existence dates make it easier to match party descriptions to the right researcher or organisation in the NLA's Trove-People and Organisations database. More information about existence dates

 

Date	Change history
14 Oct 2011	New, separate and expanded Party Best Practice page
21 Nov 2011	RIF-CS v1.3.0 change information added: Party Type (administrativePosition), NameType (superior and subordinate), using primary relationships to link all collections to a party, existence dates, using termIdentifier in collections to describe parties as subjects.
6 Dec 2011	Added links to search for NLA party ID, Scopus author ID, & ThomsonReuters ResearcherID
27 Apr 2012	Added advice that party records for international researchers may be included in Trove
4 May 2012	Added links to new Trove index page

 

 

 

 

Please send any feedback on this page to guides@ands.org.au