On Thu, Jun 17, 2010 at 3:42 PM, Karen Coyle <[email protected]> wrote: > In cleaning out my mailbox, I realize I didn't answer Tom's question > about name order... see below: > > > Quoting Tom Morris <[email protected]>: > >>> 1) cannot parse into surname/given -- it's a one-way street on that >>> account, and OL decided to take the surname, given form and put it in >>> natural order. I could use foaf:name for the natural order name. >> >> Is this decision final for all time? > > No. The "AACR" form could be added into the records at a later date, > and I am encouraging OL to consider adding the LCNA identifier for the > author as well (which then could presumably be linked to VIAF, etc.). > So it is possible that the "library form" of names will be available > at some point. (BTW, many non-librarians find the last name/first name > form for authors to be unfriendly in bib displays. It's really only > needed for sorted sets or heading browses.)
The advantage to storing the individual name pieces, whether it be in "last, first" form or, preferably, as separate fields is that you always have the option to transform to natural order for display. > This brings up an interesting issue that I'm sure others are > struggling with. Library data is really expert data -- data that you > need training and knowledge to create (semi-)correctly. Asking > untrained editors (and OL is an open wiki) to put name or title > elements into the right slots is pretty clearly too much. Not only > would there be errors in the input, but users could easily get > discouraged when faced with a lot of decisions that they would have to > make in order to contribute: What does "Title of person" mean? > "Non-filing" segment? Where do you put the name of a one-named person > like Homer? (Few non-librarians would think that "Mao, Tse-tung" is > the correct entry :-)). > > The balance between useful data and ease of input is a hard one to > find. I've mulled over in my mind some kind of bibliographic "core" (I > know, I hate to call it that) that would be expert-created, and that > would provide a stable description, with a lot of user-editable data > attached. That only works when you have expert-created data available. > And it means that you then need to manage accounts for experts who can > edit the expert data. It gets messy very quickly. > > I'd love to hear if others have approached this issue and what ideas you have. It's definitely a non-trivial problem, but I think it's exacerbated by the assumption that there will be a 1:1:1 mapping between input, storage, and display. You might try experimenting with an output only field that shows users how the name they're entering will display and/or collate. Validation routines which check for the probability of swapped names might be another crutch that would assist users. Tom _______________________________________________ Ol-tech mailing list [email protected] http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech To unsubscribe from this mailing list, send email to [email protected]
