A couple years ago, I used a crossmap of LC call numbers to subject headings (admittedly out of date) to provide subject-labeled sort by call number on an experimental catalog search.http://lisv06.colgate.edu/profound/<%20%20%20%20%20%20%3C/div%3E%09%09%09%09http://lisv06.colgate.edu/profound/%09%09%09%09%09%09%09%09%09%09%09%09> The mapping came from Mona Scott. Conversion Tables<http://encore.colgate.edu/iii/encore/search/C%7CSmona+subject+scott%7COrightresult%7CU1?lang=eng&suite=def> .<http://encore.colgate.edu/iii/encore/search/C%7CSmona+subject+scott%7COrightresult%7CU1?lang=eng&suite=def> 1999 I don't know how robust this is, but try searching a word that will appear across subject areas, like "brown", to see the classification/subject labels. I read the tables into a database, and in a batch process, coded each call number division by how deep into the hierarchy it was linked - the number of indents from 1 to 6. My ambition was to then try to find the most frequently used subject headings in each step of the hierarchy (limited to a workable range) to try to generate some semantic-net-like set of links between subject headings and classification. But I never was able to pursue that goal. Cindy Harper, Systems Librarian Colgate University Libraries [log in to unmask] 315-228-7363 On Sun, Oct 30, 2011 at 5:58 PM, David Friggens <[log in to unmask]>wrote: > > Clicking on one of Ben Shneiderman's treemapping projects reminded me > that > > I've always thought treemaps [1] would serve well as a browsing interface > > for library and archive collections because they work well with > hierarchical > > data. > > I played around with this earlier in the year, wanting to provide a > drill-down into our collections by call number. > > For our Education Library's Teaching Collection, I used a three-level > visualisation of items based on Dewey hierarchy, and coloured by the > proportion of "new" (post 2006) items. I never put it online anywhere, > so have attached it here. > > Dewey was pretty easy to get labels for the first three levels, and > that seemed reasonable enough for most areas. But the majority of our > items are LCC, and that's where I ran aground. The labels for the > first two letters are readily available, but far too general to make > this interesting. I couldn't seem to find any useful data in machine > readable format. Sourcing another level down from LoC [1] or Wikipedia > [2] seems tantalisingly close, but there's a whole lot of manual > effort in turning these (incomplete) ranges into something usable. > > Cheers > David > > [1] http://www.loc.gov/catdir/cpso/lcco/ > [2] http://en.wikipedia.org/wiki/Library_of_Congress_Classification > > > -- > oʇɐʞıɐʍ ɟo ʎʇısɹǝʌıun > uɐıɹɐɹqıן sɯǝʇsʎs >