[Web4lib] LibGuides

Varnum, Ken varnum at umich.edu
Wed Sep 30 13:47:25 EDT 2009


LibGuides offers, as part of a supplementary package (at what we felt was a modest cost), a periodic export of your library's guides.  They give you the entire content of your guide (all tabs, all content), along with author metadata, in XML.  LibGuides updates the XML file weekly.  We indexing that XML file in our site search and returning guides in several ways:

1) Individual Research Guides (our name for the content hosted through LibGuides) appear in their own section of the search results when there are keyword matches.  For example, a site search for "history" -- http://www.lib.umich.edu/mlibrary/search/mirlyn%3Bwebsite%3Bejournals%3Bsearchtools%3Bdeepblue/history - pulls up 22 guides.

2) In a topical browse, we make use of LibGuides' tags to assign individual guides to the subject category.   Users who go to our Economics browse page -- http://www.lib.umich.edu/browse/Economics - see a link to two Research Guides that have been appropriately tagged by the librarians who write them.

Having the entirety of the guide's content and metadata at our disposal in XML has proven very useful.  Since we have the data, we can (though we don't currently) do such searches as John mentions.

--
Ken Varnum
Web Systems Manager
University Library                       E: varnum at umich.edu
University of Michigan                   T: 734-615-3287
309 Harlan Hatcher Graduate Library      F: 734-647-6897
Ann Arbor, MI 48109-1205                 http://www.lib.umich.edu/





On 9/30/09 10:54 AM, "John Fereira" <jaf30 at cornell.edu> wrote:

The API, IMHO, stretches the definition of API.  It's really a form that
will produces a URL which can but cut-n-pasted onto another web page.
It's not very flexible, and doesn't expose all of the features that are
URL addressable (for example, with the right incantation you can get a
block of html which represents the tag cloud).   You can't, for example,
get a list of LibGuides which are in a named category *AND* have a
specific tag.

Still, it's worth looking at the API in terms of how you may want to
manage LibGuides for your institution.  Creating a controlled vocabulary
for Categories and Tags early on will allow you to make better use of
the API.   For example, if you maintain Subject based LibGuides and
Course based lib guides, tagging the course guides with something like
"Course Guide" (and the course number) will make it easy to pull in just
the Course Guides.  Establish a Category/Tagging policy early on and
you'll save time later.


More information about the Web4lib mailing list