Thanks - I just produced tables of circ data for FY14 and for the lifetime of the past two systems, and was thinking of 1) converting these to percentile ranks, 2) loading into our Millennium catalog records in a 59x field for webpac display and 3) using a jquery table to sort the webpac display on the ranking. Stuff to do in my off-work time.
From: Code for Libraries [mailto:[log in to unmask]] On Behalf Of Karen Coyle
Sent: Thursday, November 06, 2014 5:18 PM
To: [log in to unmask]
Subject: Re: [CODE4LIB] Wednesday afternoon reverie
Cynthia, it's been a while but I wanted to give you feedback...
Ranking on importance based on library ownership and/or circulation is something that I've seen discussed but not implemented -- mainly due to the difficulty of gathering the data from library systems. But it seems like an obvious way to rank results, IMO.
Too bad that one has to pay for BISAC headings. They tend to mirror the headings in bookstores (and ebook stores) that people might be familiar with. They capture fiction topics, especially, in a way to resonates with some users (topics like "Teen Paranormal Romance").
On 10/22/14 1:25 PM, Harper, Cynthia wrote:
> So I'm deleting all the Bisac subject headings (650_7|2bisacsh) from our ebook records - they were deemed not to be useful, especially as it would entail a for-fee indexing change to make them clickable. But I'm thinking if we someday have a discovery system, they'll be useful as a means for broader-to-narrower term browsing that won't require translation to English, as would call number ranges.
> As I watch the system slowly chunk through them, I think about how library collections and catalogs facilitate jumping to the most specific subjects, but browsing is something of an afterthought.
> What if we could set a ranking score for the "importance" of an item in browsing, based on circulation data - authors ranked by the relative circulation of all their works, same for series, latest edition of a multi-edition work given higher ranking, etc.? Then have a means to set the threshold importance value you want to look at, and browse through these general Bisac terms, or the classification? Or have a facet for "importance" threshold. I see Bisac sometimes has a broadness/narrowness facet ("overview") - wonder how consistently that's applied, enough to be useful?
> Guess those rankings would be very expensive in compute time.
> Well, back to the deletions.
> Cindy Harper
> Electronic Services and Serials Librarian
> Virginia Theological Seminary
> 3737 Seminary Road
> Alexandria VA 22304
> [log in to unmask]
[log in to unmask] http://kcoyle.net