[Koha] normalization rule - defining a matching rule

[Archives and Collections Society]

An expansion of Pete's email below; we have a major problem using 
"Normalization Rules" and "Normalization Checks" at 
<http://koha-admin/cgi-bin/koha/admin/matching-rules.pl?op=edit_matching_rule&matcher_id=1> 
to compare our biblio records with those of LoC (Library of Congress) for 
staged records.

Please note that we have searched the Koha docs; if this is a bug, please 
advise.

Take the following examples:

LoC ...   020 // $a 9004158634 (hbk. : alk. paper)
Ours ...  020 // $a 9004158634

LoC ...   020 // $a 0870213601 : $c $22.95
Ours ...  020 // $a 0870213601

Koha does *not* match these. (the paranthetical comment in the first 
example appears "legal" under MARC; the colon in the second appears to be 
contrary to MARC definitions.)

The Koha admin page above gives two possibilities: (a) defining the 
"Normalization rule:" for the matchpoint (this appears to be matcher.pm 
which looks for ISBN and the date field in 008), or (b) adding a "match 
check" which might add flexibility. We have played unsuccessfully with 
these for a few days now.

Questions:  Is there some documentation on this anywhere?  If not, has 
anyone here got a solution?

Thanks in advance,

Paul
Tired old sys-admin (now getting slightly frustrated)



At 01:57 PM 1/10/2011 -0800, you wrote:

>Can someone please point me to more info about normalization rules in the
>context of defining a matching rule in Koha?
>
>Scenario:
>Our set of books currently in Koha was imported from Marc that we created
>using MarcEdit (mapping from an Excel spreadsheet). As a result our Marc
>information is likely "not as good" as that of LOC for any given book.  So
>now we are trying to "overlay", or "overwrite" as many of the biblio records
>in our Koha database with (hopefully) "better" MARC records from LOC (or
>wherever).
>
>Problem:
>When I do a bulk z39.50 lookup on ISBN numbers (using MarcEdit), the ISBN
>numbers I receive from LOC often contain trailing text (like "paperback", or
>":").  When I stage these records and try to match the staged records on
>ISBN in Koha this trailing text prevents many matches.
>
>I am assuming that it is the job of some sort of normalization routine to
>force a match, however I am not able to find any documentation or examples
>describing how one would define his/her own normalization routine.
>
>Can someone please provide an example?
>
>Thanks,
>Pete.
>--
>View this message in context: 
>http://koha.1045719.n5.nabble.com/normalization-rule-defining-a-matching-rule-tp3335536p3335536.html
>Sent from the Koha - Discuss mailing list archive at Nabble.com.
>_______________________________________________
>Koha mailing list  http://koha-community.org
>Koha at lists.katipo.co.nz
>http://lists.katipo.co.nz/mailman/listinfo/koha

---
Archives and Collections (ACS) Society
205, Main Street, Picton, Ontario, K0K 2T0, Canada
http://www.AandC.org
Canadian Charitable Organization 88721 9921 RR0001
Dedicated to maritime conservation and education.