Currently when PTI and PCI information is stored it is linked to a Title Instance. In some cases, where some incoming data is subsequently found to be incorrect, it may be necessary to "fix" data in the local kb by re-linking the PTI and PCI data to a new Title Instance. In order to support this, we will start to store information on the PTI and PCI that can later be used to rematch PTIs and PCIs to title instances after the data for a title instance has been changed (e.g. one title instance split into two separate title instances).
The proposed information to be stored as a set of "match key" information is mainly based on data in the KBART guidelines (where this maps to a KBART field, the KBART field is given in brackets for reference):
- title string (publication_title)
- issn for print version (print_identifier for serials)
- issn for electronic version (online_identifier for serials)
- isbn for print version (print_identifier for monographs)
- isbn for electronic version (online_identifier for monographs)
- date print version published (date_monograph_published_print)
- date electronic version published (date_monograph_published_online)
- author (first author)
- editor (first editor)
- monograph volume (monograph_volume)
- edition (monograph_edition)
The first eight should be prioritised as they cover the information currently used to match incoming data to existing title instances. The remaining six could be relevant when matching monographs, but are not currently used.