Talking a bit about this on Stu Weibel’s blog, Stu invited me to send some suggestions to OCLC. Most of these suggestions can really only be done by OCLC and/or Google (although if done they’d make things possible at the end of individual libraries or other third parties). Nevertheless, after spending some time writing it up, I figured might as well share it with the internets.
1. GOOGLE BOOKS/OCLCNUM MATCHING
Google already knows about OCLCnum’s that correspond to given Google Books volumes, as evidenced by the “Find In a Library” links offered off of most Google Books volumes, where those links are based on OCLCnum.
It would be great if the Google Books API would include these OCLC numbers in the response. The old version of the API was actually documented to say it would include both OCLCnum and LCCN where available, but it sadly never actually did. The just released new version of the Google Books API doesn’t mention OCLCnum or LCCN at all.
It would be great if you could somehow get Google to include OCLCnum (and ideally LCCN too where available) in the Google Books API response. This way clients of the Google API could have an OCLCnum to link to OCLC services, as well as other library-sector services, where OCLCnum is one of the most useful identifiers.
Also, Google Books API should allow _searching_ in OCLCnum and LCCN. It actually is possible now, using an undocumented method of searching for “OCLC123455” or “LCCN12345”, but that’s not documented anywhere and it’s not clear how supported or likely to continue it us. It would be great if there were actually an officially documented way of searching for OCLCnum or LCCN on Google Books.
From the other end, it would also be awesome if WorldCat itself could include corresponding Google volume ID (or possibly multiple IDs) in WorldCat records. That would probably require Google to regularly share a file of googleID/OCLCnum correspondences with OCLC, because they probably change from time to time. But if they were in WorldCat bibs (in some appropriate marc field, although surprisingly none seems obviously appropriate to my eye; maybe 024 with an appropriate $2), and ALSO searchable via an index in worldcat APIs, that would similarly allow software to more easily connect from WorldCat services to Google services or other third party services using Google IDs.
2. WORLDCAT.ORG USER WORKFLOW
So the “Find In a Library” link from Google Books is great, but once a user clicks on it and arrives at Worldcat.org, the user experience is very sub-optimal. There are some things that could be done purely on OCLC’s end to improve things.
The user may be from my library, but my libraries holdings don’t neccesarily show up at the top of the holdings list. It depends on the user’s IP address being accurately geo-coded to a location very near my library, which not all users will be (because of inaccuracies in IP-address geocoding, and because even though they are a patron of my library they may not in fact be located closer to my library than any others!). And even if you know you want to see “Johns Hopkins Library” holdings, it’s not really clear how/if you can specifically find those.
It would be great if there were some way for a user to set a cookie-based preference expressing a particular library (or even more than one library) as their library of choice, such that once set this libraries holdings would be at the top of the holdings list. Ideally setting a library as a preference in this way would ALSO make the libraries “links” show up — the links including OpenURL links that show up now only if the user is from a recognized IP address belonging to a library. Ideally this preference could be expressed _without_ the user needing to create a WorldCat account, just in an ephemeral cookie — having to create an account is too much of a barrier for user experience. Although if the user DID choose to create an account, they should be able to set this sort of library preference attached to it too.
We still have a problem with access/authorization though. It is my understanding that users from IP addresses associated with my library see MORE holdings in worldcat.org than the general public (holdings from other libraries than mine, which even my users will sometimes be interested in), becuase my library is a FirstSearch WorldCat subscriber. However, most of my users use is NOT from an on-campus IP address — and if, from off campus, they follow a “Find In a Library” link to worldcat.org, they won’t get this expanded access, and there’s no way I can think of for them to get it. It would be great if worldcat.org would allow a shibboleth-based login to FirstSearch WorldCat customer libraries enterprise authentication, so a user could choose to authenticate as a Johns Hopkins user by clicking on an appropriately labelled link, and get expanded access. Ideally such Shibboleth functionality would allow WAYFless shibboleth login when the user IS following a link from my library.
Another issue is that once a user DOES find a holding listed for my library and click on it, the experience is still sub-optimal. That banner/frameset that WorldCat insists on putting on the link is really nasty, interfering with at least the design if not sometimes the functionality of local OPAC/ILSs, and generally making the user’s experience seem clunkier. I really wish that could go away.
And additionally, more powerful direct linking templates could be supported by Worldcat for linking to a local OPAC. Worldcat allows an ISBN template and OCLCnum template, and chooses which one to use based on the data available and the templates available. However, our new open-source based OPAC/discovery layer would also allow links with the meaning “ISBN:X OR OCLC:Y OR LCCN:Z”, which would probably maximize the chances of a succesful link, and it would be great if WorldCat would support such linking.
3. MORE OPTIONS FOR GOOGLE BOOKS OUTGOING LINKS
This last one may not be in the business interests of either OCLC or Google, but I’ll include it anyway.
The “Find In a Library” link is a good first step, and if the user experience on OCLC’s end could be improved could end up pretty great.
However, as someone at an individual library, it is unfortunate for the library ecosystem in general that the ONLY way from Google Books to an individual library’s services runs through worldcat.org. It would be good if there were a way to get additional individual library-specific links on a Google Books page. One model for this is the OpenURL linking provided by Google Scholar, where a user can choose institutional preferences on Google Scholar preferences, and have institution-specific OpenURL links added to pages. In fact, the Google Scholar OpenURL preferences could conceivably be used on Google Books too, the same preferences set could result in OpenURL links to the same destinations being added to Google Books results too. If these OpenURL links included ISBNs, OCLCnums, and LCCNs where available, they’d be pretty useful incoming links to a library.