Presentation is loading. Please wait.

Presentation is loading. Please wait.

Introduction to Persistent Identifiers

Similar presentations


Presentation on theme: "Introduction to Persistent Identifiers"— Presentation transcript:

1 Introduction to Persistent Identifiers
Summary Definition of persistent identifiers Which objects – physical / digital Proliferation of identifiers What do identifiers allow us to do The challenge of real persistence Social contracts Present, real examples Course on persistent identifiers, Madrid (Spain) Type of Persistent Identifiers Kevin Richards Software Developer Landcare Research February 8th, 2012

2 Introduction This presentation explains some characteristics of a typical identifier scheme, some examples of types of identifiers and briefly looks at some issues that may arise when the information associated with an identifier changes This session is presented by Kevin Richards, Software Developer, at Landcare Research in New Zealand.

3 Summary Types of Persistent Identifiers Characteristics of an ideal identifier scheme Commonly used identifier types Points for consideration Events that impact the maintenance of identifiers Discussion

4 Characteristics of an Identifier scheme
Universally unique Unchanging Independent generation Opaque Associated services

5 Characteristics of an Identifier
Universally Unique A mechanism is required to ensure global uniqueness Integers obviously not good here Internet example – domain names, URI UUID ensures uniqueness with no tie to resolution protocol

6 Characteristics of an Identifier
Unchanging The resource the identifier refers to must remain the same The link between the identifier itself and what it identifies is the key point Draw diagram to illustrate – eg ID – refers to this object, etc

7 Characteristics of an Identifier Scheme
Independent generation, ideally It should be possible to create identifiers without relying on a central service If a generation mechanism has been defined to ensure uniqueness, then the creation of identifiers can be done by anyone, anywhere c.f. DOI We are mobilising millions of objects

8 Characteristics of an Identifier Scheme
Opaque, ideally In theory, should not be possible to determine any information about a resource by looking at the identifier Often not easy to achieve May be more of a sociological issue

9 Characteristics of an Identifier
Associated Services Not required for a persistent identifier but makes the identifier more useful Resolution of identifier Other functions may include metadata requests and generation Explain uselessness of a UUID by itself

10 Resolution of Identifiers
URLs, URIs, URN E.g. http, lsid, mailto, ftp HTTP Protocol for resolving a web resource which may have many formats, e.g. image, html page, text, xml May have redirects, HTTP 303 Resolution proxies

11 Types of Identifier Schemes
UUID (sometimes GUID) Assured unique E.g. d e1-b86c c9a66 Hard to type in Not resolvable Not always DB friendly Opaque

12 Types of Identifier Schemes
HTTP URI Uniform Resource Identifier Web based – uses HTTP and DNS In common use Promoted by Linked Data advocates May not be opaque due to semantics of domain names E.g. A generalisation of URLs.

13 Types of Identifiers PURL Persistent Uniform Resource Locator
Eg Web based using HTTP and HTTP redirect Resolved through PURL resolver May not be opaque due to domain names and paths

14 Types of Identifiers DOI Digital Object Identifier
Eg doi: /182 Managed by DOI Foundation (commercial) Generated by DOI Foundation Resolved through DOI resolution service Very opaque

15 Types of Identifiers LSID Life Science Identifier (URN)
Developed by OMG for “name” based identification E.g. urn:lsid:example.org:specimen:12921 Resolution protocol independent (i.e. does not rely upon HTTP) 3 step resolution mechanism May not be opaque due to domain names

16 When things change Change of institution name (domain name for URLs)
E.g. old ID new ID Broken Identifiers ? Avoid by avoiding institution names for authority, use Project names. E.g use NZspecimens.org.nz instead

17 When things change Transfer of dataset (or split)
New owner for dataset Eg specimens of NHM given permanently to regional herbarium -> new institution name Similar solution to last problem, but if project name changes too, then still an issue, e.g. NZspecimens.org.nz project is migrated to wellingtonSpecimens.org.nz 2 options new owner maintains old Identifiers Old Identifiers redirected to new identifiers

18 When things change Resource destroyed
When the resource the identifier refers to has been deleted or removed from the system Preferred results is to maintain the resolution of the identifier and state it is deprecated in the metadata for that resource

19 Summary/conclusions Characteristics of an Identifier
Universally unique, independent generation, unchanging, opaque and actionable Types of Identifiers (some of them) URI, PURL, DOI, LSID When things change Split or transfer of dataset, change of institution name, destroyed resource

20 Introduction to Persistent Identifiers
Summary Definition of persistent identifiers Which objects – physical / digital Proliferation of identifiers What do identifiers allow us to do The challenge of real persistence Social contracts Present, real examples Course on persistent identifiers, Madrid (Spain) Type of Persistent Identifiers Kevin Richards Software Developer Landcare Research February 8th, 2012


Download ppt "Introduction to Persistent Identifiers"

Similar presentations


Ads by Google