>> The following FRBR manifestation attributes have been ignored:
>
>FYI, much of this I don't care about. However, I probably do care
>about publication status (if that's covering the MODS' notion of
>"issuance" -- which includes "continuing" as in a serial, and
>"monographic") and numbering (though not quite sure what this is!).
I'm not sure that's what "publication status" is - my memory recalls
that loosely as a "yes, it's still being published", "no, it's dead",
or "unknown at this time" equivalent. Refer to the FRBR PDF for a
more technical description (and if my recollection is wrong,
certainly correct me).
Are you talking about "frequency" of an issue, like yearly, or
monthly, that sorta thing? In FRBR, this is defined at the
expression level, and is something also ignored in the
existing data model:
The following FRBR expression attributes have been ignored:
context, sequencing pattern (serial), expected regularity
of issue (serial), expected frequency of issue (serial) ...
The only real date work incorporated into the data model now
is libdb_events, which covers corporate bodies and their
meetings and so forth (from the data model notes on CB):
FRBR defines a "numerical designation sequencing a meeting,
conference, exhibition, fair, etc. that constitutes one of a
series of related meetings, conferences, exhibitions, fairs,
etc., or any other numerical designation associated with a
corporate body." These would be implemented as "events"
and associated with a "relationship".
Reference the libdb_events table for more information.
Briefly, it contains three date-related columns, thusly:
start_date: The date (YYYY-MM-DD) the event started.
end_date: The date (YYYY-MM-DD) the event ended.
frequency: How often the event occurs (e.g., yearly,
monthly, weekly, biannually, etc.).
As such, the following statements could
be made in the existing data model:
SerialA is type "serial", described in a WEMI model.
SerialA has relationship "publication status", which
maps to an entry (EventSerialsA) in libdb_events.
EventsSerialA defines a start_date of when the serial
started publishing, an end_date (possibly blank) of
when it ended, and a frequency of "quarterly", etc.
This, of course, is all off-the-top-of-my-head thinking - I've tried
to make the data model as flexible and "pure FRBR" as possible,
but I've only been thinking of it from an implementation standpoint
regarding movies/films, etc., etc. So, the above could all have
fatal flaws. I dunno yet. Haven't really thought about it.
I also haven't thought about handling of serials from an "issue"
standpoint - I have no current comments on reproducing "Vol. 12
Iss. 13", "Vol. I, Issue VIII" or "December 2004, Holiday Special"
within the existing data model.
--
Morbus Iff ( you, me, eropuri? aawwwwwWWWw yYeahahhHHAhhh )
Culture: http://www.disobey.com/ and http://www.gamegrene.com/
Spidering Hacks: http://amazon.com/exec/obidos/ASIN/0596005776/disobeycom
icq: 2927491 / aim: akaMorbus / yahoo: morbus_iff / jabber.org: morbus
|
