Enterprise Abstraction

Content is king, long live content!

Some thoughts (none particularly original) I'm having about library futures.

In the library world, I keep seeing posts from libraries and librarians all wondering about how to be relevant, valued, important.  Many talk about the importance of librarianship, and the specialized knowledge and skills they have.  We focus on better ways to search, and licensing content.  Indeed these things have value.  But I'm not sure they're actually the core enduring value for libraries.  I think the decrease in the perceived value of libraries is more fundamental, and simpler.

It's about content.  Libraries used to have content that was unavailable, or otherwise very expensive to own.  Sharing via the library was a sensible and logical thing to do.  Today that still makes sense - though the types and volume of information that's available via the internet/web reduces the value for many kinds of information (encyclopedias for example, and also books and journals that are available in digital formats - ie:  all of them practically).

By my thinking though, the key purpose for a library - providing access to information that is generally otherwise unavailable - is still an enduring value.  It's simply that the types of information that are generally unavailable has changed.  Today thats not books and journals.  It is never-published and/or un-digitized materials, scientific data, lab notebooks, photographs, rare books and manuscripts, old maps, records, documents of enduring (national, organizational, historic) value.  Our value is in identifying, preserving, and making these materials available.

Does that mean that libraries should become archives?

CISTI Mobile

Wow, it's been a long time since I've posted. 

In that time, we've launched CISTI Mobile in production.  In general, the creation was a positive experience. We learned plenty about JQuery mobile, Metalib API's (used to enable the federated search), and about mobile website development generally. 

I think, correct me if I'm wrong, that is was indeed the first Canadian federal library mobile website?

Having accomplished the first goal of providing for a mobile interface for CISTI, we'd now like to pursue the further goal of having people actively use it.  We've a few daily users.. but there's plenty of room for more usage. 

I'd been thinking of redirecting mobile users from our various applications to the mobile site - but now I realize that that's kind of complicated.  The simplest solutions to implement wouldn't place them on their equivalent landing page, and thus would likely cause confusion.  It seems that this is one of the key disadvantages to having a "separate" mobile site, rather than skinning existing applications to serve mobile users natively.

That said, the effort to skin all of our applications would have been enormous.  Hence, I think I'll see if we can't simply direct users to the common landing pages/static pages to the mobile site - and leave the rest for now.

 

 

List of Canadian Repositories

I've done some recent experimentation involving harvesting Canadian institutional repositories with the OAI-PMH protocol.  Mostly this was a personal endeavour to determine how many there were, how much content they had, and how difficult/reasonable it would be to harvest them all and provide a search index.  It seems that there's no comprehensive listing of Canadian repositories, so I had to gather and cobble my own. 

It seems not to be a common practice for repositories to publish their OAI-PMH addresses.  This is a minor tragedy.  The list that I gathered follows for the edification of others who might have a similar need.

The list consists of 53 unique Canadian repositories, and where available, their OAI-PMH interface API's.

An experimental harvest (future availability uncertain) of these repositories resulted in over 367093 unique metadata records.  The harvester used was an instance of the PKP Open Archives Harvester.

Updated July 4, 2012 to add Colonial Despatches

 

Repository	OAI-PMH Base URL
Archipel	http://www.archipel.uqam.ca/perl/oai2
Atrium	http://atrium.lib.uoguelph.ca/oai/request
AUSpace DEV 1.5	http://auspace.athabascau.ca:8080/oai/request
Avian Conservation and Ecology	http://www.ace-eco.org/oai/
Brock University Digital Repository	http://dr.library.brocku.ca/oai/request
Canadian Breast Cancer Alliance	https://researchspace.library.utoronto.ca/oai/request
Canadian Bulletin of Medical History	http://www.cbmh.ca/index.php/cbmh/oai
cIRcle - UBC's Information Repository	http://circle.ubc.ca/oai/request
Collection mémoires et thèses de l'Université Laval	http://www.theses.ulaval.ca:8080/oaicat/servlet/OAIHandler
Colonial Despatches	http://bcgenesis.uvic.ca/oai.xq
Constellation	http://constellation.uqac.ca/cgi/oai2
Curve	http://curve.carleton.ca/oai
DalSpace at Dalhousie University	http://dalspace.library.dal.ca:8080/oai/request
Dépôt numérique de UQTR	http://depot-e.uqtr.ca/cgi/oai2
Digital Commons @ McMaster	http://digitalcommons.mcmaster.ca/do/oai/
Digital Commons @ Ryerson	http://digitalcommons.ryerson.ca/do/oai/
DSpace at the University of Lethbridge	http://www.uleth.ca/oai/request
[email protected]	http://dspace.ucalgary.ca/dspace-oai/request
eCommons::Research	http://ecommons.uwinnipeg.ca/oai/request
eCommons@USASK	http://ecommons.usask.ca/dspace-oai/request
ERA: Education and Research Archive	http://era.library.ualberta.ca/oaiprovider/
Érudit	https://depot.erudit.org/dspace-oai/request
eScholarship@McGill	http://digitool.library.mcgill.ca/OAI-PUB
Espace ETS	http://espace.etsmtl.ca/cgi/oai2-etdms
International Developmental Research Center (IDRC)	https://idl-bnc.idrc.ca/dspace-oai/request
International Review of Research in Open and Distance Learning	http://www.irrodl.org/index.php/irrodl/oai
IslandScholar	UNKNOWN
LU Zone UL	https://zone.biblio.laurentian.ca/dspace-oai/request
Memorial University Research Repository	http://research.library.mun.ca/cgi/oai2
MSpace at the University of Manitoba	https://mspace.lib.umanitoba.ca/dspace-oai/request
MSVU Digital Commons	http://dc.msvu.ca:8080/dspace-oai/request
MUN Digital Archive Initiative	http://collections.mun.ca/cgi-bin/oai.exe
NRC Publications Archive	http://oai.cisti-icist.nrc-cnrc.gc.ca/oai/ctrl
oURspace	http://ourspace.uregina.ca/dspace-oai/request
OZone	http://ozone.scholarsportal.info/dspace-oai/request
Papyrus - Dépôt institutionnel numérique	http://papyrus.bib.umontreal.ca/dspace-oai/request
Polar Data Catalogue	http://www.polardata.ca/oai/provider
POLYPublie	http://publications.polymtl.ca/cgi/oai2/
PubMed Central Canada	UNKNOWN
Qspace	http://qspace.library.queensu.ca/dspace-oai/request
Recherche uO Research	http://www.ruor.uottawa.ca/dspace-oai/request
RiverRun @ UNB	http://dspace.hil.unb.ca:8080/oai/request
Scholarship@Western	http://ir.lib.uwo.ca/cgi/oai2.cgi
Sémaphore	http://semaphore.uqar.ca/cgi/oai2
SFU Institutional Repository	http://ir.lib.sfu.ca/dspace-oai/request
Spectrum: Concordia University Res. Repository	http://spectrum.library.concordia.ca/cgi/oai2
Theses Canada	http://oai.collectionscanada.gc.ca/oai/oai.php
T-Space at The University of Toronto Libraries	http://tspace.library.utoronto.ca/tspace-oai/request
Université du Québec à Chicoutimi - Documentation régionale	http://sdeir.uqac.ca/metadata/oai.asp
University of Saskatchewan Electronic Thesis and Dissertation Archive	http://library.usask.ca/etd/OAI/oai.pl
UVicSpace	http://dspace.library.uvic.ca:8080/oai/request
UWSpace	http://www.uwspace.uwaterloo.ca/dspace-oai/request
WinSpace	UNKNOWN
York Space	https://www.library.yorku.ca/dspace-oai/request

Journal cover images - in SFX Link Resolver

I been learning SFX administration - adding some features here and there... last week I added the Google Books covers service - which turned out to be quite satisfactory.  It got me thinking that there must be a similar service for journal covers.  Apparently not.

However, some colleagues in the Twitterverse recommended using the journal covers supplied by Elsivier.  It's not an API call, nor am I able to directly call the images remotely - but I have some space - so I decided to download and link them locally.

All the files are named XXXXXXXX.gif - where XXXXXXXX is the ISSN of the journal.  (I've since renamed the files to XXXX-XXXX.gif to support the way that SFX deals with ISBN's).  After downloading and untarring the image files, I placed them in a web-accessible directory.

Then, it was simply a matter of linking, unobtrusively, to the images in SFX (or presumably from any application that deals with ISSN's).  For SFX, I inserted the following - but this code would work for any page as long as you can insert the desired ISSN in the appropriate places:

<!-- SKA EXPERIMENTAL - Adding journal cover images -->
<!-- Will show image if available, or will show 0 height 1px image otherwise -->
<!-- Note:  Without Javascript in IE, will show no-image icon if image does not exist -->
   <TMPL_IF ISSN>
        <a href="http://lab.cisti-icist.nrc-cnrc.gc.ca/anthonys/journalcover/<TMPL_VAR NAME="ISSN" ESCAPE='HTML'>.gif">
          <img src="http://myserver.mydomain.ca/journalcovers/<TMPL_VAR NAME="ISSN" ESCAPE='HTML'>.gif" height=90 alt="" onerror="this.src='http://myserver.mydomain.ca/journalcovers/px.gif'; this.height='0';">
        </a>
    </TMPL_IF>
<!-- END SKA EXPERIMENTAL -->

In my case, I inserted this code in sfxmenu.tmpl, just after the area where the Google Books cover would load.  The <TMPL_IF ISSN> prevents this portion from being rendered if the item in question does not have an identified ISSN (for example - if it were a book, or other work).

One could certainly that the portions - removing the <TMPL> portions that are specific to SFX, and insert this code anywhere.  In most cases, if the ISSN image is not available, it will display px.gif (which is a 1 pixel white image).  In Firefox it's totally transparent - but in IE, it will show a broken image if the ISSN image is not found - and the user has disabled JavaScript.  I hope to resolve that problem eventually.

Some examples:

Example 1

Example 2

Example 3 (No cover available)

Change

We've recently released a new website for CISTI - reorganizing some things in conjunction with the CISTI Transformation and the changes in document delivery services.

Externally, the website has changed it's structure - and is still transitioning as the rollout of the new partnership for document delivery proceeds.  That said, essentially the same services and content remain available.  We've added a new metasearch function - that we hope will simplify things - particularly hoping to simplify the answer to the question "Does CISTI have this?"   Now you should be able to find out - in one place.

Internally, things have changed significantly.  In the past, CISTI provided a separate portal dubbed the "Virtual Library" for the exclusive use of NRC.   We have merged the functionality of this portal with our main website - http://cisti-icist.nrc-cnrc.gc.ca - such that IP recognized persons (or those using our NRC proxy server) will see many more options including the ability to browse and search licensed and free e-journals, databases and other resources, librarian authored subject guides, and have an assortment of federated search options.  Integrating the ILS/Catalogue, metasearch, and CISTI-hosted databases (Discover and NPArC) is CISTI's link resolution service Find@CISTI - now finally being used essentially EVERYWHERE there's a bibliographic citation presented on the website.

Presently, we're working to improve the site, and to release (finally) a version of the LibX Toolbar for NRC Staff - that I hope will bring the capabilities of the library more readily available for researchers in their work.

It's been a busy time.

SFX and JavaScript

I'd been hunting around for a means to have our local SFX implementation (Find@CISTI) run without requiring JavaScript.  I'd done the requisite Google search, and submitted a request - to no avail.  The solution turned out to be rather trivial - and seems to have been implemented in subsequent releases of the SFX templates.

In your SFX instance, at /templates/sfxmenu/target_list.tmpl, add the following noscript line:

<!-- end extra comments --> <tr> <td height="0"> <noscript><input title="Navigate to target in new window" type="submit" value="<TMPL_VAR NAME='SERVICE_NAME'>&nbsp; <TMPL_VAR ESCAPE=HTML NAME='TARGET_NAME'>"></noscript> </td> </tr>

Our line had been removed or was missing... problem solved.  The buttons presented with JavaScript off are not particularly pretty - but they are functional. 

OpenURL and COInS at CISTI

CISTI has just released OpenURL links and COInS in CISTI Discover and NPArc (the NRC Publications Archive, NRC's Institutional Repository).

OpenURL in CISTI Discover and NPArc permit recognized users to connect with their libraries link resolver to get a list of appropriate full text content and other information services.  In the case of NRC staff, they will be recognized and directed to the NRC-CISTI link resolver to be directed to electronic content held or licensed on their behalf, or alternately to the means to search for and procure that content.  Others (not NRC) are presently directed to a default link resolver that provides a wide variety of choices and options. 

COInS (ContextObjects in SPANS) help applications, such as the browser plug-ins LibX and Zotero, to recognize the citation information and to provide users options such as saving/bookmarking the references.

Pics of the CISTI Stacks

For the second time in my life I actually visited CISTI's stacks today - the areas on the 4th floor and above where CISTI keeps the dead trees.  I snapped a few pics for my personal enjoyment that others might like.  It's hard to get a feel for how large the space actually is, partly because it's divided over 4 floors, but also because there's no place on any of the floors to get a full pic of the entire area due to the building design.  Anyway, I thought some others might be interested, hence the set is on flickr.

I particularly liked this item, carefully shelved amongst the extensive collection of scientific works.  I may have to borrow it:

 

NRC Publications Archive (NPArC) Launched

CISTI has launched the NRC Publications Archive (NPArC) providing access to a large portion of the research output of the National Research Council.  Already there is a significant set of documents available.  From the notice:

"NPArC (pronounced N-Park) houses research papers for refereed journals, peer-reviewed conferences and technical reports that NRC researchers produce each year. It also guarantees long-term access to this research — serving as a valuable resource for NRC researchers, collaborators and the public. NPArC currently includes 35,000 publication records from 17 NRC institutes, along with 5,000 full-text publications from four of these institutes. And, over the course of this year, CISTI will be translating all new abstracts into both official languages."

To support the initiative, NRC has established a policy effective January 2009 making it mandatory to deposit copies of all peer-reviewed, NRC-authored publications and technical reports in NPARC.  The NRC License to Publish (Crown Copyright) form has been updated to support this requirement.

The NPArC web interface supports searching by a wide variety of criteria, and also supports browsing by subject, author or NRC Institute.

For those interested, NPArC metadata is also harvestable via OAI-PMH at:  https://accs.cisti-icist.nrc-cnrc.gc.ca/oai/ctrl?verb=Identify

NPArC is based upon CISTI's Trusted Digital Repository infrastructure that I've mentioned previously.

Making allies of adversaries in libraries

[This post concerns mainly libraries serving a campus of patrons - public libraries or those with a general public mission have other concerns]

Google is not the enemy.  Nor publishers, nor vendors (entirely - but I wouldn't turn my back on a vendor in a dark alley). 

The library role based on having/holding content is diminishing.  More and more libraries (particularly science libraries) no longer hold collections but instead license content. 

Moreover, the role of a library as a search enabler is also compromised by technologies.  WebOPACs and ILS systems, federated search platforms are diminished by the fact that they are either not client-oriented interfaces (WebOPAC), or that if others don't the job better - they at least are further up on the list of patrons first stops on their research journey (Google, Google Scholar, Scopus, etc).

So, what to do?  Turn your adversaries into allies.

Libraries will not be a starting place for researchers - deal with it.

Libraries will not be the holders of information - the value of "collections" is diminished - deal with it. 

How?  Libraries still have a valuable role - to be connectors to information.

Indeed, encourage patrons to search anywhere/everywhere!  Go to Amazon if that meets your needs, browse wikipedia, start at your favourite journals website, use Scopus and other tools if you have access!  I believe that it is the job of the library to get patrons to information - regardless if that information is in our collection or not.  And regardless of where the patron decides to begin their search.

It is possible, with a small set of commodity technologies to bring access to patrons no matter where they start.  In order of importance:

1. Link resolver - connects your own technologies and 3rd parties to appropriate copies and information services for your patrons.
2. Proxy server - allow off-campus access to IP-restricted resources.  Alternately, you might use Shibboleth or some other kind of authentication/authorization mechanism.
3. Federated search - Allows for broad searches across local, remote, paper and digital collection.  Not a silver bullet - but at least your patrons are not expected to perform 30 searches on 30 different data sources - and still remain generally unaware of the vast quantity of expensive resources you license on their behalf.  More importantly perhaps, a federated search platform can be the foundation for your "discovery layer" - hooks and landing areas to connect multiple information services and sources together.  In a rare instance, a patron might even begin their search there - but I doubt it.  Finally - at the very least - the patron will not be forced to look at your WebOPAC... nobody should have to look at that.  Apply all the lipstick you want.
4. Library Toolbar - You can make one for your library via LibX in an hour.  This is for me the paradigm shifter.  It is well known that patrons rarely start their search at the library.  Don't try to change patron behaviour, go with it!  If they start their search on a 3rd party website - the library toolbar should travel with them on their research journey - pointing out places where the library can help them.  Oh, I see a ISBN, or a citation, or a DOI - here's a link back to your library to find out more information or get more services!   
5. Other goodies - Alerting, citation management, visualization, analysis tools, recommendations, facets, personalization... all of these and more can be tacked onto the federated search and link resolver... you're adding more and more value!

Why are we afraid of Google and others?  Don't fear them, embrace them!    And connect patrons from wherever they go to where they need to be.