task #8792
Updated by Andreas Müller over 4 years ago
The functioning of the ERMS import (and PESI export) is explained at: [[Erms_mapping]]
Many issues have been handled in #1444. However as the ERMS-CDM is complex in often difficult there are many issues to take care for during upcoming imports.
* Misapplications
Currently misapplications are identified by a UNION statement in ErmsTaxonImport.getAcceptedTaxaKeys(). It searches for misidentification patterns in tu.tu_unacceptreason . We also tried to evaluate the tu_authority field but searching for "sensu", "non", "auct". But realized that all these patterns are not used specific enough in ERMS. "Auct" is often simply used for misspellings, "sensu" is used for unassessed taxa, genus transfer, sensu in the meaning of sec and others or simply unclear. "Non" is generally to unspecific as it can be used for homonyms, too.
So only the above query is possible but discussion with VLIZ is needed on data cleaning or standardization. Maybe even an own tu_status for misapplications is possible.
* Multiple open issues to be hanled by or discussed with VLIZ, mentioned in email to VLIZ 2019-12-23
* tu_acctaxon IS NULL
~~~ sql
SELECT *
FROM tu
WHERE tu_acctaxon IS NULL
~~~
should be checked. VLIZ should fix these (n=479). Currently they are handled as accepted taxa.