Matching rules on record import
Hi there I'm trying tu get my Koha3 up and running, but need some directions setting Matching Rules to import MARC records. First of all ... Matching rules seems to be working ok ... I've used the ISBN rule and got good results.... BUT ... there are some problems when the records don't have ISBN data (or ISSN) ... so I need a way to match those. for example .. I have this record (already imported into Koha) and when I try to import again it has "no match" (of course, because it has no ISBN) LDR 00645nam#a2200253#a#4500 001 CON000008 003 OUN 005 20060601105400 008 901227| xxu |||| ||eng d 040 _aOUN 080 _a3-055.2 080 _a341.16:362.7 092 _sNU _c3-055.2 _lH431u 100 1 _aHazzard, Virginia. 245 10 _aUnicef and women : _bthe long voyage : a historical perspective 260 _aNew York : _bUnicef, _cc1987. 300 _a148 p. ; _c27 cm. 440 _aUnicef history series. _pMonograph ; _v7 650 04 _aUNICEF 650 04 _aMUJERES 866 0 _a1 ej. 907 _aTEXTO 930 _a90476 _dREST _pv 942 _cBK So ... I need advice on how to set up a valid matching rule to handle a situation like this one. The records will come from various libraries (around 13) each one with distinctive control numbers (field 001) so I think maybe comparing this field and title (field 245$a) would give me a good enough matching rule. the problem is ... how to set it up on Koha :) Frist thing I don't understand completely ... when creating a new matching rule I must specify a "search index" value (on ISBN rule it has 'isbn'). Now, for something like what I need, what should be the value, if it has to be one... Thanks in advance -- Ing. J. Martin Longo Depto. Gestion Informatica SID - UNCuyo http://sid.uncu.edu.ar
Hi, On Tue, Feb 17, 2009 at 9:29 AM, Martin Longo <jmlongo@uncu.edu.ar> wrote:
The records will come from various libraries (around 13) each one with distinctive control numbers (field 001) so I think maybe comparing this field and title (field 245$a) would give me a good enough matching rule.
Yes, if you know that the source control numbers are unique, the 001 would be a good match point.
Frist thing I don't understand completely ... when creating a new matching rule I must specify a "search index" value (on ISBN rule it has 'isbn'). Now, for something like what I need, what should be the value, if it has to be one...
Use the Control-number index, which in the MARC21 indexing definitions indexes the 001 field. Regards, Galen -- Galen Charlton VP, Research & Development, LibLime galen.charlton@liblime.com p: 1-888-564-2457 x709 skype: gmcharlt
The records will come from various libraries (around 13) each one with distinctive control numbers (field 001) so I think maybe comparing
I'd like to jump in this conversation, too, because I haven't been able to get the match rule to work and got only one response last time I queried the list. I'm trying to match on the OCLC control number, which seems to be in both 001 and 035. I tried matching on 001, as Galen advised below, and using "control-number" as the index, and I've also tried using "ocm" as the index. Here's what I have now: Description: OCLC control number Threshold: 1000 Search index: control-number Score: 1000 Tag: 001 Subfields: a Length: 0 Offset: 0 Normalization rule ?? I added a match check with the same information (field 001, subfield a). Can anyone see what I'm doing wrong? Karen Kohn Collection Development Manager Landman Library Arcadia University 450 S. Easton Road Glenside, PA 19038 ph: 215-572-8528 fax: 215-572-0240 -----Original Message----- From: koha-bounces@lists.katipo.co.nz [mailto:koha-bounces@lists.katipo.co.nz] On Behalf Of Galen Charlton Sent: Tuesday, February 17, 2009 9:54 AM To: Martin Longo Cc: koha@lists.katipo.co.nz Subject: Re: [Koha] Matching rules on record import Hi, On Tue, Feb 17, 2009 at 9:29 AM, Martin Longo <jmlongo@uncu.edu.ar> wrote: this
field and title (field 245$a) would give me a good enough matching rule.
Yes, if you know that the source control numbers are unique, the 001 would be a good match point.
Frist thing I don't understand completely ... when creating a new matching rule I must specify a "search index" value (on ISBN rule it has 'isbn'). Now, for something like what I need, what should be the value, if it has to be one...
Use the Control-number index, which in the MARC21 indexing definitions indexes the 001 field. Regards, Galen -- Galen Charlton VP, Research & Development, LibLime galen.charlton@liblime.com p: 1-888-564-2457 x709 skype: gmcharlt _______________________________________________ Koha mailing list Koha@lists.katipo.co.nz http://lists.katipo.co.nz/mailman/listinfo/koha
Hi, On Tue, Feb 17, 2009 at 2:10 PM, Kohn, Karen <kohnk@arcadia.edu> wrote:
Tag: 001 Subfields: a
Try leaving the subfield field blank in the import rule - the 001 field doesn't have subfields. Regards, Galen -- Galen Charlton VP, Research & Development, LibLime galen.charlton@liblime.com p: 1-888-564-2457 x709 skype: gmcharlt
Hi, I'd like to jump in this conversation, too, to know if it is possible to match on the 001 field in noZebra mode. I tried to create an index on this field and I could not find any book with a search on this index. This is a sample of the window NoZebraIndexes in /Administration/System preferences/Cataloging/ 'lccn' => '001', 'control-number' => '001', However, it is impossible to match on the 001 field. Thanks for help. G. Hatt Kohn, Karen a écrit :
I'd like to jump in this conversation, too, because I haven't been able to get the match rule to work and got only one response last time I queried the list.
I'm trying to match on the OCLC control number, which seems to be in both 001 and 035. I tried matching on 001, as Galen advised below, and using "control-number" as the index, and I've also tried using "ocm" as the index.
Here's what I have now:
Description: OCLC control number Threshold: 1000 Search index: control-number Score: 1000 Tag: 001 Subfields: a Length: 0 Offset: 0 Normalization rule ??
I added a match check with the same information (field 001, subfield a).
Can anyone see what I'm doing wrong?
Karen Kohn Collection Development Manager Landman Library Arcadia University 450 S. Easton Road Glenside, PA 19038 ph: 215-572-8528 fax: 215-572-0240
-----Original Message----- From: koha-bounces@lists.katipo.co.nz [mailto:koha-bounces@lists.katipo.co.nz] On Behalf Of Galen Charlton Sent: Tuesday, February 17, 2009 9:54 AM To: Martin Longo Cc: koha@lists.katipo.co.nz Subject: Re: [Koha] Matching rules on record import
Hi,
The records will come from various libraries (around 13) each one with distinctive control numbers (field 001) so I think maybe comparing
On Tue, Feb 17, 2009 at 9:29 AM, Martin Longo <jmlongo@uncu.edu.ar> wrote: this
field and title (field 245$a) would give me a good enough matching rule.
Yes, if you know that the source control numbers are unique, the 001 would be a good match point.
Frist thing I don't understand completely ... when creating a new matching rule I must specify a "search index" value (on ISBN rule it has 'isbn'). Now, for something like what I need, what should be the value, if it has to be one...
Use the Control-number index, which in the MARC21 indexing definitions indexes the 001 field.
Regards,
Galen
-- ======================================================== Guillaume HATT Bibliothécaire Informatique documentaire Ecole nationale des chartes 19 rue de la Sorbonne 75005 Paris Courriel : guillaume.hatt@enc.sorbonne.fr Tél. : 01 55 42 75 05 ========================================================
We added a 245a matchpoint for 020 and 022. The matchpoint was modeled on the example. Tag: 245 Subfields: a Offset: 0 Length: 0 Normalization rule: TITLE We have one for 035 too that seems to work okay. I added a 245a matchpoint for it today though I'm not sure that it is necessary. Tag: 035 Subfields: a Offset: 0 Length: 0 Normalization rule: OCLC https://sites.google.com/a/liblime.com/koha-manual/Home/Table-of-Contents/ad... Rachel Hollis, librarian Stevens-Henager College, Boise Idaho Campus -----Original Message----- From: koha-bounces@lists.katipo.co.nz [mailto:koha-bounces@lists.katipo.co.nz] On Behalf Of Galen Charlton Sent: Tuesday, February 17, 2009 7:54 AM Subject: Re: [Koha] Matching rules on record import On Tue, Feb 17, 2009 at 9:29 AM, Martin Longo <jmlongo@uncu.edu.ar> wrote:
Frist thing I don't understand completely ... when creating a new matching rule I must specify a "search index" value (on ISBN rule it has 'isbn'). Now, for something like what I need, what should be the value, if it has to be one...
Use the Control-number index, which in the MARC21 indexing definitions indexes the 001 field. Regards, Galen This message is for the designated recipient only and may contain privileged, proprietary, or otherwise private information. If you have received it in error, please notify the sender immediately and delete the original. Any other use of the email by you is prohibited.
Am I correct in assuming that Koha can only use one matchpoint at a time? We've recently completed an inventory of our collection and we ended up with several different indentifiers for our items: ISBNs, LCCNs, & OCLC numbers. It would be nice if during the import process Koha could check a sequence of matchpoints when checking for duplicate holdings (e.g., first check ISBN, then LCCN, etc.) This is currently not possible, however, right? Thank you, Cab Vinton, Director Sanbornton Public Library Sanbornton, NH
Hi, On Thu, Feb 19, 2009 at 5:23 AM, Cab Vinton <bibliwho@gmail.com> wrote:
It would be nice if during the import process Koha could check a sequence of matchpoints when checking for duplicate holdings (e.g., first check ISBN, then LCCN, etc.) This is currently not possible, however, right?
You can define multiple matchpoints in an import matching rule. To do this, you start by setting a match threshold. The specific value doesn't matter, but 1000 is a good default. You can then add more multiple match points and assign a numeric score to each. For example, suppose your rule has a threshold of 1000, and you want to consider LCCN, ISBN, and biblios.net record number. The biblios.net ID is stable, so you could trust that it's a good matchpoint and give it a score of 1000 - which means that if there's a match on biblios.net ID, it's enough to identify a record to overlay. Let's say you don't trust the ISBN or LCCN as much, and give each a score of 500. That means that a record doesn't match on biblios.net ID, it would have to match on both ISBN and LCCN to reach the threshold of 1000. Regards, Galen -- Galen Charlton VP, Research & Development, LibLime galen.charlton@liblime.com p: 1-888-564-2457 x709 skype: gmcharlt
Thanks for the responses. It is just not working, and I am stuck. I set mine up like Rachel Hollis's: Search index: ocm Score: 1000 1. Tag: 035 2. Subfields: a 3. Offset: 0 4. Length: 0 5. Normalization rule: oclc The match check repeats the same information. Is that what it's supposed to do? Any other ideas? Karen Kohn Collection Development Manager Landman Library Arcadia University 450 S. Easton Road Glenside, PA 19038 ph: 215-572-8528 fax: 215-572-0240 -----Original Message----- From: koha-bounces@lists.katipo.co.nz [mailto:koha-bounces@lists.katipo.co.nz] On Behalf Of Rachel Hollis Sent: Wednesday, February 18, 2009 7:54 PM To: 'Galen Charlton'; Martin Longo Cc: koha@lists.katipo.co.nz Subject: Re: [Koha] Matching rules on record import We added a 245a matchpoint for 020 and 022. The matchpoint was modeled on the example. Tag: 245 Subfields: a Offset: 0 Length: 0 Normalization rule: TITLE We have one for 035 too that seems to work okay. I added a 245a matchpoint for it today though I'm not sure that it is necessary. Tag: 035 Subfields: a Offset: 0 Length: 0 Normalization rule: OCLC https://sites.google.com/a/liblime.com/koha-manual/Home/Table-of-Content s/administration/Manual--Catalog/Catalog--Record-Matching-Rules Rachel Hollis, librarian Stevens-Henager College, Boise Idaho Campus -----Original Message----- From: koha-bounces@lists.katipo.co.nz [mailto:koha-bounces@lists.katipo.co.nz] On Behalf Of Galen Charlton Sent: Tuesday, February 17, 2009 7:54 AM Subject: Re: [Koha] Matching rules on record import On Tue, Feb 17, 2009 at 9:29 AM, Martin Longo <jmlongo@uncu.edu.ar> wrote:
Frist thing I don't understand completely ... when creating a new matching rule I must specify a "search index" value (on ISBN rule it has 'isbn'). Now, for something like what I need, what should be the value, if it has to be one...
Use the Control-number index, which in the MARC21 indexing definitions indexes the 001 field. Regards, Galen This message is for the designated recipient only and may contain privileged, proprietary, or otherwise private information. If you have received it in error, please notify the sender immediately and delete the original. Any other use of the email by you is prohibited. _______________________________________________ Koha mailing list Koha@lists.katipo.co.nz http://lists.katipo.co.nz/mailman/listinfo/koha
participants (6)
-
Cab Vinton -
Galen Charlton -
Guillaume Hatt -
Kohn, Karen -
Martin Longo -
Rachel Hollis