Upgrade StorageMore sophisticated choice of generic translators
Currently Zotero uses the generic translators in the order unAPI, COinS, DOI, Embedded Metadata
As more and more sites embed high quality google highwire metadata, I wonder if we can come up with something better. On almost all occasions, google metatags are preferable to both COinS and DOI. Well implemented, they give users not just abstracts and tags, but even attached PDFs (which the others never do).
I realize, of course, why the current order of priority is the way it is, but I wonder if we can't come up with something smarter. One simple solution would be to put google, bepress (and soon eprints) into a separate translator with higher priority (i.e. a lower number) but maybe there are other solutions?
I thought I throw this out for discussion before opening a ticket.
As more and more sites embed high quality google highwire metadata, I wonder if we can come up with something better. On almost all occasions, google metatags are preferable to both COinS and DOI. Well implemented, they give users not just abstracts and tags, but even attached PDFs (which the others never do).
I realize, of course, why the current order of priority is the way it is, but I wonder if we can't come up with something smarter. One simple solution would be to put google, bepress (and soon eprints) into a separate translator with higher priority (i.e. a lower number) but maybe there are other solutions?
I thought I throw this out for discussion before opening a ticket.
-
aurimasWe could use DOI/COinS/unAPI to do the initial scrapping and then supplement the metadata with EM. The only issue there is that DOI/COinS/unAPI could be referring to a different item than EM would, so we would need to do some checking to make sure we're merging correct metadata.