New Terms - bibliographicIdentifier & bibliographicIdentifierType

[nyctalus-noctula]

New term

• Submitter: Jacek Stefaniak
• Efficacy Justification (why is this term necessary?): Information provided by dcterms:bibliographicCitation, especially when citation format is not specified, can be too vague to pinpoint individual reference. Even if some type of URI is provided in citation, it is not easily accessible for machine reading or linking to exterior database. This proposition aims to provide unmistakable ID to bibliographic resource with flexibility to user which catalog system they want to use. Additionally, linking reference to specialized bibliographic database allows to access more detailed information about resource, not provided by Bibliographic Resource class (for example: language, date, overall number of pages, series, journal title etc.), more in line with best practices and standards in academic citations
• Demand Justification (name at least two organizations that independently need this term): Proposition is a result of DwC-DP workshop organized by Polish GBIF Node, attended by representatives of 10 academic institutions
• Stability Justification (what concerns are there that this might affect existing implementations?): None
• Implications for dwciri: namespace (does this change affect a dwciri term version)?: None

Proposed attributes of the new term:

• Term name (in lowerCamelCase for properties, UpperCamelCase for classes): bibliographicIdentifier
• Term label (English, not normative): Bibliographic Identifier
• Organized in Class (e.g., Occurrence, Event, Location, Taxon): Bibliographic Resource
• Definition of the term (normative): Number or symbol to uniquely identify a book or document
• Usage comments (recommendations regarding content, etc., not normative): Recommended best practice is to use a globally unique identifier issued by recognizable authority (i.e. International ISBN Agency, International DOI Foundation)
• Examples (not normative): 978-0565095024; 10998406; 10.1046/j.1420-9101.1997.10010039.x
• Refines (identifier of the broader term this term refines; normative): http://purl.org/dc/terms/bibliographicCitation
• Replaces (identifier of the existing term that would be deprecated and replaced by this term; normative): N/A
• ABCD 2.06 (XPATH of the equivalent term in ABCD or EFG; not normative): N/A

Proposed attributes of the new term:

• Term name (in lowerCamelCase for properties, UpperCamelCase for classes): bibliographicIdentifierType
• Term label (English, not normative): Bibliographic Identifier Type
• Organized in Class (e.g., Occurrence, Event, Location, Taxon): Bibliographic Resource
• Definition of the term (normative): Type of persistent identifier
• Usage comments (recommendations regarding content, etc., not normative): Recommended best practice is to use a limited, tightly controlled vocabulary of identifier issuers.
• Examples (not normative): ISBN; DOI; ISSN; OCN
• Refines (identifier of the broader term this term refines; normative): http://purl.org/dc/terms/bibliographicCitation
• Replaces (identifier of the existing term that would be deprecated and replaced by this term; normative): N/A
• ABCD 2.06 (XPATH of the equivalent term in ABCD or EFG; not normative): N/A

[tucotuco]

@nyctalus-noctula The term referenceID, proposed here, subsumes the term bibliographicIdentifier you are proposing. Similarly, bibliographicIdentifierType is covered by referenceType, proposed here.

[nyctalus-noctula]

As refrenceType states, it is:

A category that best matches the nature of a dcterms:BibliographicResource

So rather type of object which consist reference (book, book section, journal article, oral communication). I propose type of URI, issued by some sort of global catalogue authority to access additional information on resource.

If I understand refrenceType wrongly, it should be better described to avoid misinterpretation.

Bibtex format was inspiration for this issue, and it states type of resource as existing type of object (article, book, conference) alongside separate ID in form of doi or crossref. So, IMHO those 2 types of information shouldn't be mixed in one refrenceType field.

[tucotuco]

@nyctalus-noctula I have added some suggested examples in the issue for referenceType.

As for the URI type, wouldn't that be abundantly obvious from the URI itself?

[nyctalus-noctula]

@tucotuco thanks for acknowledging issue.

As for the URI type, wouldn't that be abundantly obvious from the URI itself?

In case of DOI yes, but I think that we can specify it for extra clarity - sometimes ISBN and ISSN can be mixed up, and if someone use more obscure URI it will help to find issuing authority.