getting technical: reference linking (or, “well, i wouldn’t start from here…”) tom bishop...
TRANSCRIPT
Getting technical: reference linking(or, “Well, I wouldn’t start from here…”)
Tom Bishop
Information Services Manager,The Royal College of Surgeons of England Library
3rd June, 2003
Getting technical
On the agenda (possibly)...
Getting technical
DOI
CrossRef
OpenURLSFX
Balsa
OPACs
1CATE
SWALK
Blink 182P45
EIEIO
The key players
How do we get access to journals and citations?
• Publishers – their own web sites and services• Full text aggregators – Journals@Ovid, ProQuest, etc.• ‘Hosts’ – SwetsWise, Ebscohost, HighWire, etc.• ‘Secondary’ (A&I) databases – PubMed, Cambridge Scientific Abstracts, etc.• … and libraries!
Getting technical
Reference linking: what do we mean?
“The ability to go directly from a citation to the work cited” (Priscilla Caplan)
• Links among and between journal articles• Links from bibliographic citations to journal articles• Seamless, easy access allowing the user to link quickly and directly to any document they want
Getting technical
…And you may ask yourself - “Well, how did I get here?”
“Internal linking”• Links contained within one service - e.g. from a reference in a full-text aggregator’s service to the text of an article held in the same service.
“External linking”• Takes the user out of a service to somewhere else - e.g. from a database to a publisher’s website, from one publisher’s website to another’s, etc.
Getting technical
CrossRef
External linking initially largely dependent on bilateral agreements.
CrossRef is:
• “…a collaborative linking service”
• a membership network, not a product
Getting technical
CrossRef
Over 200 publishers registered
Agents - HighWire, IngentaAffiliates - Ebsco, TDNet, SwetsBlackwellLibraries
7.8 million content items registered across over 7000 journals
How does it work?
Getting technical
Take a step back…
A link:
http://www.rcseng.ac.uk/services/library
• ‘Static’ – link is computed in advance
• Not necessarily persistent - relates to a specific location defined at a specific time
Getting a bit more technical
DOI: Digital Object Identifier
A unique identifier assigned to a digital object.
• A way of accessing an object (e.g. a full text article) without having to know its URL - the DOI identifies the object itself, not the place where it is stored.
• Persistent - as long as the object exists, so does the DOI.
Getting even more technical
DOI: Dumb Old Identifier
“Dumb” number - doesn’t relate to the object, couldn’t be guessed - like a phone number.
10.1000/123456789
Getting even more technical
Prefix - given to the rights owner e.g. the publisher
Suffix - any unique alphanumeric string
e.g. 10.1074/jbc.M004545200
CrossRef and DOI
• Publisher participating in CrossRef creates unique DOI for article.
• Associates it with:• the articles’s metadata e.g. journal title/ISSN, vol., issue, pages, title, author, AND• the URL where it resides.
• Submits all this to CrossRef for registration in a central directory.
Getting even more technical
CrossRef and DOI: the benefits
• Participants can link in to the article by retrieving from CrossRef the DOI that links to it.• Publishers can link outward from an article’s citations to any content registered by CrossRef.• Single agreement rather than many bilateral ones.• Drives up journal traffic.• If the location of an article changes, all the publisher has to do is update the URL in the CrossRef directory.• If the journal changes publisher, the DOI stays the same.
Getting a bit less technical, hopefully
…But along comes the librarian with his bucket of cold water...
• Many articles do not have a DOI, and there are publishers not involved in the scheme.
• “With CrossRef it only takes a click or two to get to the full text, either as an authenticated user or through pay-per-view services.”
The cat/pigeons slide
Is this good enough??
Work versus manifestation
A CrossRef DOI represents the work (the authoritative version of the article as published) itself, but not manifestations (different versions) of that work.e.g. the print/PDF/HTML versions may differ.
Partly soluble in this example - publisher’s page to which the DOI for that article resolves can point to different versions.
But then we hit…
Getting a bit more technical again
The dictatorship of the publishariat
Van der Sompel: “Dictated linking”
• ‘Target’ (i.e. where the user ends up) is dictated by the link provider.
• Ability to link and range of targets depends on the link provider’s collections or agreements.
• Bypasses the local environment - the link provider tells the user where to go, and the local institution cannot influence that.
Why is this important?
That’s not a real word,is it?
The issue of ‘appropriate copy’
CrossRef DOIs resolve to a publisher’s ‘response page’ - usually a full bibliographic citation, often an abstract, and a mechanism to get to the full text. But the user will usually be denied access to full text if they don’t have a personal or institutional subscription. Pay-per-view may be offered but…
Getting a bit more technical again
…the user may in fact be able to get hold of the article from a different source as a result of their library affiliation.
Appropriate sources and services
• Another database, aggregator or host to which the library subscribes• The library’s print holdings• Document delivery/ILL - the library’s own or a specific provider
Solution = ‘Context-sensitive linking’
• Also allows for ‘Extended services’ - internet search, link from OPAC book record to book review on the web, etc.
Getting a bit more technical again
SFX (Ex Libris)
First product to attempt to solve these problems.
How does it work?
Link source: the information resource where the user begins, e.g. a citation in a database or within an article, or a journal record in an OPAC.
Link target: the information resource where the user ends up, e.g. publisher response page, TOC page, journal home page etc.
Getting even more technical
Hang on…
… How does the link resolver ‘know’ about the article required?
Rather than a static link to a particular target, the link source provides a ‘hook’ that gives information, in the form of metadata, about the article.
When a user clicks the ‘SFX’ button, an OpenURL is created.
Getting even more technical
OpenURL
A way of transmitting metadata in a URL - essentially a transport system.
http://LinkFinderPlus.library.edu?genre=article&issn=12345678&volume=99&issue=1&date=20020101&spage=27&atitle=What Is An OpenURL&title=Harry's White Papers
As technical as it gets
‘Base URL’: web address of the link resolver i.e. the address to which the Open URL is being sent
‘Query’: the metadata that the link resolver uses to identify and link to appropriate targets
OK, but…
… How does the information resource provider in whose service the user has found the citation (e.g. database provider, journal publisher, host, OPAC, etc.) ‘know’ the affiliation of the user and therefore which link resolver to send the OpenURL to?
• OpenURL-aware sources (e.g. ScienceDirect, SwetsWise, Ovid databases) - link resolver address provided in institutional profile.• Use of ‘cookie pushers’. • Local solutions like IP recognition.
As technical as it gets
OpenURL and CrossRef/DOI
Complement each other when used in conjunction.
http://LinkFinderPlus.library.edu?genre=article&issn=12345678&volume=99&issue=1&date=20020101&spage=27&atitle=What Is An OpenURL&title=Harry's White Papers&id=doi:123/456789
• Link resolver can point the user to a persistent object as an appropriate copy.• DOI is attached to metadata in the CrossRef database – no need to update a local link resolver if things change.
As technical as it gets
Everybody’s doing it…
LinkFinderPlus (Endeavor)
1CATE (Openly Informatics)
WebBridge (Innovative)
LinkSource (Ebsco)
Linksolver (Ovid)
Balsa (EDINA)
Getting less technical again
Open URLs and context-sensitive resolution - a good example
Andy PowellOpenResolver: a simple OpenURL resolver.
Articlehttp://www.ariadne.ac.uk/issue28/resolver/
Walkthroughhttp://www.ukoln.ac.uk/distributed- systems/openurl/
Getting less technical again
The benefits...• Appropriate copy and extended services - decided at a local level to maximise usage.
• Better user interface - users look for SFX (or whatever) button wherever they are, cuts out wild goose chases.
• Links aren’t computed until the user clicks – dynamic “just-in-time” linking reduces response times.
• ‘Political’ - puts the library at the centre of searching and retrieval.
Not too technical
…and the costs?
• Financial - dependent on product and institution
• Time and effort
“The SFX services… are only as good as the information provided by the librarian. The more detailed the information, the better the level of granularity for linking.” (Jenny Walker, SFX)
“An OpenURL resolver is never better than the knowledge base behind it.” (Eric Hellman, Openly)
Not too technical
Future developments
• National or centralised resolvers?
• Multiple resolution of DOI
Technical for the last time
“I believe OpenURL resolution is going to become as obvious a necessity for navigating through and across databases as online catalogs were for libraries in the 80's.” (Chuck Hamaker, UNCC)