Harvest references from attachments.

It would be nice for zotero to capture a reference list from attachments (notably pdfs). I guess the list could be in a tab on the attachment property sheet. Even nicer if zotero could check if each item in the list is in the local zotero database, and provide a link to it. And (I'm getting breathless) offering the opportunity to add/download selected items that don't exist locally would be marvellous.

Mendeley offers a nice implementation of reference harvesting (better than its automatic metadata creation).
  • I agree that this would be really neat, but in my experience this sort of thing ends up producing such deeply flawed data that while at first glance looks impressive and flashy tends to just stick a bunch of junk in your database. It could be a neat plugin for someone to work on.
  • Junk data's a possibility, but I suspect that's equally so with the pdf metadata extraction that Zotero currently has. Keeping a zotero server-based store for tuning might help (I think Mendeley does this with its metadata extraction).
  • Not so, since our PDF metadata association relies on data that's canonical, at least in principle. It's not parsed directly out of PDFs.
  • Fair enough. It could be useful though even if imperfect. The harvested references could be automatically cleaned up by comparing with a canonical source, where available, perhaps?
Sign In or Register to comment.