Recently I sat in on a presentation regarding Ex Libris’s newest service, bX, which is basically a “find similar” application that ties together articles across platforms.
From my notes:
- focus is on article as core unit of use
- use is correlated over the course of a particular browser session (with a timeout and negative weighting for reference desk stations)
- based on structural analysis, not just popularity, so if A and B are co-used, then they show as related
- recommendations are hyperlinked only if the article is available full-text from your library (and can make it so only full text articles show in the first place)
- embedded in SFX (although it can be used by others) and harvests OpenURL data
Although we probably aren’t going to attempt to implement it while we’re getting MetaLib and Primo up and running, I’m hoping we can find the money later to do so. It looks like a very handy tool.
And if you’re interested in some of the research behind this, see the research projects Usage-based measurements of journal quality and MESUR.
