[Web4lib] Interesting Web/Library 2.0 data(wasparticpationSkillsfor Library 2.0 Leaders)

[John Fereira]

At 10:34 AM 5/9/2007, Lars Aronsson wrote:
>Walker, David wrote:
>
> > A system like that would not only be as easy to use as Google,
> > it would be in many ways better, since it would be tailored to
> > our community,
>
>So cut down on the words and show me a URL where this has been
>proven.

This isn't specifically for the library community but I developed 
this search interface for a system that I built in partnership with 
the USDA and the library I work in (Albert R. Mann Library, Cornell 
University).  First take a look at http://www.usda.gov and their 
search box.  That system uses the Google Search Appliance to index 
all of the content in the usda.gov domain (it actually indexes some 
content in our library as well).   Try entering "title:poultry" in 
the search box.

Now take a look at http://usda.mannlib.cornell.edu   Enter "title: 
poultry" in the search box here.  The result list will contain a list 
of documents which have "poultry" in the title.  Essentially it shows 
that this search form will accept the same syntax as one would use 
for searching Google.  The difference is how the result are presented 
and more importantly how the content to be searched is indexed.  For 
example, try "title:corn agency:nass".  Note that now the results are 
limited to reports with poultry in the title from the NASS (National 
Agricultural Statistics Service" (BTW, if you try the same search on 
the usda.gov site you won't get any hits).  Try clicking on the 
"Poultry Slaughter" title.  Because the result set from the search 
contains a unique report identifier It can be used to look up lots of 
additional metadata and build a page which provides access to all of 
the issues of Poultry Slaughter.  Check out the "Subject Headings" 
box on the right.  That shows all of the subject headings in which 
"Poultry Slaughter" was classified, thus providing a mechanism to 
discover other titles with the same classification.

The search mechanism was built using the Java Lucene API.  It allows 
us to index content based on custom fields (i.e. column names in the 
backend database) , customize the results however we want, but still 
use the Google search syntax, something users are familiar with.







>--
>   Lars Aronsson (lars at aronsson.se)
>   Aronsson Datateknik - http://aronsson.se
>_______________________________________________
>Web4lib mailing list
>Web4lib at webjunction.org
>http://lists.webjunction.org/web4lib/

John Fereira
jaf30 at cornell.edu
Ithaca, NY