On Sun, Feb 13, 2011 at 11:27 AM, Karen Coyle <kco...@kcoyle.net> wrote: > Quoting Tom Morris <tfmor...@gmail.com>: > >> Thanks for the quick reply, Karen. Are you saying that the >> OpenLibrary policy is to follow what's dictated in the Library of >> Congress authority file? There would seem to be at least some >> divergence in that they use a different name order, include >> birth/death years (sometimes), etc. > > No, I'm not saying that OL follows LCNA -- you asked for a source to > make decisions and the only ones I know of in the area of books are > library name authority files. Those files are based on some sensible > rules (IMO) so they might be usable as guidance. At the same time, > they do some ugly things, like put names in last,first order, which > looks very artificial.
Freebase also has some sensible rules, but they're different from the Library of Congress ones. I could pick one of the two or just make up some of my own, but, as a collaborative effort, it would seem like OpenLibrary should have its own set of rules. Also, trying to devine the LC rules based on a sampling of evidence is unreliable. For example, Freebase treats "Editors of Time," "Time staff," etc as aliases for the corporate author "Time." The Library of Congress has a smattering of entries for "Editors of <foo>" but not enough to cover the number of times they must have received books cataloged with it, so I'm left guessing as to what their policy is. Having said that, I don't really care what they're policy is unless it's the one OpenLibrary follows. >> The nice thing about computers is that they're not limited to having a >> single name at the top of the 3x5 card. If the record has both Samuel >> Clemens and Mark Twain, the computer can find it under either one. If >> they are two separate records, there needs to be a way to link the two >> (which I don't think OpenLibrary has, does it?). > > Any author record can have alternate names, and many do. However, OL > doesn't currently replicate the "see also" structure that is in > library authority data. I suspect it would be easy enough to add the > structure, a bit more tricky to fill in the data. If it is added, it > would be nice to see it displayed as: "Also writes as..." and have the > relationship recorded in both/all related records. Just for the record, the Freebase policy on this is to treat all aliases equally, so pseudonyms, maiden names, corporate name changes, etc are all treated equally. The biggest weakness with this is shared pseudonyms where multiple authors used the same name, either sequentially or as a collective. >> If the policies aren't set in stone yet though, I'd personally argue >> for something that's a little more approachable and common sense >> based. > > There are two schools of thought on that: 1) some rules are good 2) > rules make participation more difficult. Some of us like having rules, > some find them constraining. I think this will be a constant struggle. Perhaps instead of rules, a set of guidelines or a style guide is needed, if the word "rules" sounds too confining. As a collaborative effort, I don't see how OpenLibrary can avoid having at least some minimalist framework to promote consistency. Tom _______________________________________________ Ol-discuss mailing list Ol-discuss@archive.org http://mail.archive.org/cgi-bin/mailman/listinfo/ol-discuss To unsubscribe from this mailing list, send email to ol-discuss-unsubscr...@archive.org
