Getting the Goods: Libraries and the Last Mile

In my continuing series of publishing my Access 2006 notes, Roy Tennant's keynote on finishing the task of connecting our users to the information they need is something to which every librarian should pay attention.

If you don't understand something I've written, there's always the podcast of Roy's talk. In fact, there are a ton of podcasts of individual Access 2006 talks available from the Access 2006 Speakers, Presentations and Podcasts page. It's the next best thing to actually being there...

As always, any errors in capturing Roy's thoughts are undoubtedly mine.

Roy Tennant

The Last Mile

• analogous to the telecom problem of having an amazing network but being unable to make the connection to the customer
• we've done a better job describing things than we have delivering them
• our systems suck for online delivery
• after retrospective conversion and implementing automation, we've basically stalled and are being overtaken by other players like Google
• We need the 'two-click solution' (one click to find, one click to get full-text)

Books

raw Google search for "before taliban" -- first hit links to the full-text (successful two-click find-get solution) Melvyl forces users to click locations/details, then scroll down, then click through a copyright notice

• Why are libraries so bad at this? Partially because we have been asking for the wrong things from our vendors

But Google Books search doesn't provide a link to the full-text version in the top search hits at all Mirlyn (UMich) catalogue recently began working with Google to digitize 100,000 books; so let's see how things are going: Search for croquis and italiens Publication date - 1890 Compare the results in Mirlyn:

• Brief display links to the physical version, when of course the users in all probability would prefer the electronic version
• Clicking on details shows 2 electronic results - full text from UMich, but and the Google Books version which provides only the "snippet view"; if you search within Google Books then you still don't get full text but you will get segments of the book

OCLC WorldCat is working on "Audience Levels" to indicate what level of audience a given book is appropriate for (on a range of "Schooler" to "Scholar") xISBN service implements some level of FRBR to provide related ISBNs and enables NCSU's catalog availability service to find other items that satisfy the user's request So people are going to be coming through Google Books and Open Worldcat, we need to accept this and work well with those services to direct users to libraries when appropriate Currently, libraries can provide better discovery of open content than Google is, because Google Books is mostly about publishers However, libraries need to fix their discovery systems, because they're broken -- this is going to take time

Link resolvers:

• We tend to throw intermediate levels of windows at users as a search result
• Why don't we just give them full text immediately if the resolver knows that it's there? Why the heck do we link to ILL if full-text is available?
• Let's recognize that the OpenURL resolver is a beginning, not the end

Umlaut at Georgia Tech:

• Checks for full text, checks for on-shelf availability, checks for closest search results, and surfaces all of those with the bib record
• Read more about the Umlaut

GUF: Getting users to full-text

Check for David Lindahl's slides for a deeper description Try to get to full-text

• Improve metadata transfer
• Eliminate errors
• Eliminate clicks

If you can't get users to full-text, get them to print

Journals: Key Components

• OpenURL resolver
• OpenURL resolver directory
• Other resolvers (DOI)
• Aggregators (arXive.org, oaister.org)

Berkeley sends professors an email saying "we found that you published this article, but it's not freely available in our IR, please click this link to upload a copy!" and this seems to be a successful strategy to building a repository. Make it as easy as possible for the users to contribute.