On 2017-09-25 03:31 PM, Christopher Davis wrote:
Dear Colleagues,
I know that Koha can match the ISBNs of imported MARC records with the ISBNs of existing records, but is there a way to configure a record matching rule to try to match more than one match point at a time (for example try to match ISBN, then try ISSN, then try LCCN, then try OCLC number, then try standard number, then try publisher number)?
Have you tried ==> Administration ==> Record Matching Rules? We have defined seven rules, mostly the obvious ones e.g. ISBN, ISSN (we don't use OCLC -- can't afford them -- and I'm not sure what you call 'standard number'), author, title, publisher, pub date and manufacture date) and then refined them with dates from 008 etc. It takes a little time to fine tune the flexibility. It gets a very few false positives "Are you sure this isn't a duplicate?" but otherwise appears to be extremely reliable. We dropped LCCN matching -- many of our older books (we're basically an international archival library) don't have one, and the regex "treatment" of leading spaces, hyphens and suffixes became CPU cycle-heavy for minimum joy. YMMV and HTH. Best -- P.