jump to navigation

SWORD reverberations January 9, 2009

Posted by Mia in Frontiers, Repositories, SWORD.

As things progress with the IR, some of the overarching issues are welling up, from SWORD deposit to managing (or at least, trying to manage) name identification (aka authority control).

As archives and IRs steadily (albeit slowly) grow in content, there is potential for a mind-boggling number of versioning issues, in addition to the name variations issue. If multiple copies are deposited into additional archives/repositories via SWORD (potentially, since there aren’t many full-fledged implementations as far as I can glean), this seems to potentially complicate retrieval results it seems to me.

With anything like multiple-deposit, we end up with, well, multiples being deposited, along with multiples of their variations/versions – so there are many sources of location. We then have to put mechanisms in place to identify/distinguish among them in some intelligent fashion.

Maybe we don’t distinguish per se, but rather show a cascade of ‘work-like’ objects along a timeline, or some such thing. It could look more like a continuum, where the ‘work-like’ thing itself morphs through various phases:

  • the earliest version [e.g., those laying claim to some scientific results first made public]
  • subsequent edited versions
  • the most recent version [last mod date version]
  • the most authoritative version [say, one with important data corrections, annotations amended to it]
  • the published version [DOI and OpenURL hooks]
  • pdf-final of published version [green copy with publisher attribution]

But maybe those phases of an individual’s output are simply integral parts of someone else’s research process which one slogs through, the equivalent of paying-one’s-research-dues.

The thinking aloud continues.