[Koha] importing, marc mapping, progress

[Steven F. Baljkas]

Sunday, August 7, 2004     00:59 CDT

Greetings, all,

from Subject: Re: [Koha] importing, marc mapping, progress


> * Steven F. Baljkas <baljkas at mb.sympatico.ca> wrote:
> > > It seems to work during searches, but is never displayed.  Is
> > > this a template problem?
> >
> > This is one of the things I was concerned about: how Koha
> > handles the data from MARC records. Stephen, Joshua, other
> > cognoscenti: is Scott's problem here because of a mapping
> > error? something that needs to be defined in MARC parameters?
>
> I've noticed a few other fields now which are not displayed.
> Some of them show up in Koha's marc view of a record, but others
> don't even show there.  I can get them to display by going
> through the marc structure and setting the tab option, though.
> Most are set to "ignore" by default, but I think I'd prefer to
> just set each Nxx tag to use tab N.
>
> BTW, modifying the template slightly allowed me to display a bit
> more info, such as the series title.  I'm still trying to get the
> subtitle and additional authors to work, among others.  (subtitle
> doesn't show up, and only the first additional author is shown)

I still have to finish re-reading Stephen Hedge's answer to another e-mail I
sent -- it finally made its way back only yestereve (I don't know what the
daemon is doing Stephen) -- but the problems are bigger than I'd thought.

Apparently, only one subject field can be indexed (in the sense that we mean
in the library field) and that is a significant issue.

> I'd like to dump the rest of the marc data onto the book display
> page, too, but I haven't done so yet.

That would be great, Scott.

> > 1. I'm not 100% certain what you meant by "bibliographic data about
> > authors", Scott.
>
> I suppose I'll have to implement an "author detail" page and
> appropriate database storage for it.  The detail I'm looking for
> is probably not appropriate in a marc record.

I think that's for the best. Otherwise, the record would get pretty big.
Having thought more about it since I last wrote you, I think it might be
wise to designate a 59x field of your own choice for that data. That way,
if/when you eventually share data with other libraries, all that neat
information you'd be providing could be easily excluded by more
traditionalist cataloguers, without loss of other more standard information
(i.e. if you used a 500 or 520 for both the "author detail" and regular
notes, a global search and replace would end up killing some data everyone
would want).

> > 2. "... short book review blurbs ..."
> >
> > That's easily a 520 $a, first indicator "1"  (Review).
>
> I see that now, mapped already to biblio.abstract.  However, it
> is marked as non-repeatable.  Is it safe to change that?

Okay, this is the core of the problem that -- if I understood what Stephen
H. wrote me -- is freaking me out. If I did understand correctly, Koha
doesn't really understand or make use of the data in repeated fields, only
the first instance of the field per record. That is a huge problem for
proper subject cataloguing, but it also makes a problem for notes. It isn't
necessarily common to have several 520s in a given record, but it certainly
is allowed by MARC standards. (If you look at the MARC Concise Bibliographic
information on the LC website, Scott, you can see that next to 520 -
Summary, Etc. is the R indicating the field can repeat; NR = Not
Repeatable).

I gather that you can change the value but not what Koha really is doing
with the data, which would be to ignore the 2nd or further instances of a
520.

> > On another note (pardon pun), someone was asking me recently
> > whether, if you did a search for a particular work, say by
> > title, whether the author (and other, e.g. subject, possibly
> > title, fields) would come up in Koha as 'hyperlinked', if there
> > were other works by that author. This is something that Voyager
> > does and it is very handy for patrons and staff alike.
>
> Koha seems to do this somewhat already, but I hope to extend it
> to more fields.

Perhaps because I am basing my knowledge on the testdrive version available
commonly, I could be wrong here, but Koha doesn't actually seem to do what I
meant. Then again, even LC doesn't have that feature activated in their
Voyager system, although local college and university libraries near me have
those linking features available in their Voyager and Dynix systems.

> > 9. " ... quotes from books/authors/series ..."
> >
> > I can't quite see the point of going this far, Scott, but you could
handle
> > it through 500, 520 (if from the book) or 5xx notes, or even by an 856
> > linking to quotation pages.
>
> I see it as similar to why imdb.com includes quotes, reviews, and
> even discussion about their movie listings.  It helps the viewer
> decide if the data is valuable or interesting.  It may be better
> to just link to external sites for this info, though.  I could
> work out something for amazon or iblist or similar.

I guess, from my experience, Scott, it wouldn't be the library's job to try
to sell a person on a book or movie, like the commercial sites you
reference. I can see the value in the information they provide, but given
the realities of cataloguing, it is more likely that the technical services
staff would continue in a more traditional vein. If you look at the rather
detailed records I sent you off listserv for comparison sake, the dna-856
set demonstrate more what tech services staff would be inclined to do if
allowed (for everyone else, these records included some embedded
authorities, appropriate 856 notes, 246 notes for common misspellings in the
titles that could be expected, as well as complete analysis -- 505, 700
name-title and 740 -- of the contents of anthologies present in the sample).

>
> > 10. "and a few others."
> >
> > If you want to write in with the rest, or send them to me off
> > listserv, I'll be glad to tackle them, Scott.
>
> I'm unsure how to handle subtitles.  I'm not sure how Koha's
> "title" and "unititle" differ.  It sounds like one would list the
> full title, and the other would do the same without "the" and
> such prepended.  Is there a way to map biblio.title to 245a + 245b?

Again, if I've understood correctly, "title" would equal 245a but "unititle"
would stand for Uniform Title and would be a 240.

Tag 240 doesn't occur very often, really. The only time I used it was when I
had to catalogue a few different versions of the Bible for inclusion in a
collection. Otherwise, you'll see it for particularly prolific authors -- 
like our favourite and archetype, Shakespeare -- where the text per se is
more important than the edition.

Tag 245 should list the full title of a work as well as its statement of
responsibility (authorship, editorship, etc.). The second indicator in both
240 and 245 indicates how many characters the computer should skip in
processing the title that follows: initial definite and indefinite articles
plus one space (for English anyway) give the value in the second indicator.

"The same without 'the'" would be the filing title, and could be coded in
the 852l (ell) if one wanted. Normally, the computer should be able to do
the calculation from the 245 indicator and delete the appropriate number of
characters before producing a title list.

> Also, part numbers are a bit confusing.  If I have book 3 in a
> series, should that be a 245n field, 440v, 440n, or something
> else?  The series title seems to be 440a.

You had misunderstood the meaning of 245n in your original go at this,
Scott. The number and named parts there are for aspects of the title, things
that distinguish it from other works with the same 245a (and possibly 245b).
They don't come up very often in most cataloguing of books; more with legal
and report type stuff.

What you were aiming for is the establishment of a series title, which is in
the 440v in your specific case.
   E.g.     440  /0 $aHitchhiker's guide to the galaxy ;$vv.3.
would be an appropriate series entry for the work entitled: Life, the
universe, and everything.

> BTW, thank you very much for all the help!  It would take me
> forever to learn all the intricacies of marc's thousand tags.  :)

No prob. It takes a while to learn them but you're making great progress.
And it won't take forever. Don't panic! ;-)

Cheers,
Steven F. Baljkas
library tech at large
Koha neophyte
Winnipeg, MB, Canada


---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.732 / Virus Database: 486 - Release Date: 29/07/2004