Over at TechnoBiblio, Steph mentioned that the Web needs some weeding, due to the prevalence of old information and broken links. She goes on to talk about the ability to conduct searches for recently updated content. Having read the post, I found myself wondering what the weeding of an online reference collection might entail.

For example, let's say that you maintain both a general collection of subject-specific links and a physical collection covering the same topics. It seems to me that the motivations and approaches to weeding these collections might be quite different.

When I weed my collection, one significant consideration is whether the book is ever used (please note public library bias). Well, what's the online equivalent...a click-through? Would I eliminate the links that no one ever clicks on? (And at the risk of betraying my ignorance, is this easily measurable?)

What about space considerations? We often weed to make room for newer materials. In a physical space, it's both a practical and aesthetic issue. Do the same constraints apply to an online link collection? Or is an online collection's size determined mostly by the librarianly ideals of selectivity balanced with breadth and/or depth?

I, of course, have no answers...only more questions.

I wonder how lii.org deals with the issue of weeding its ever-growing collection. I'm sure they know which links are broken, but do they know which resources aren't used? Does it matter, based on their objectives? Probably not, as I would imagine that the more classic reference source evaluation factors take center stage: currency, authority, accuracy, and all of your other grad school favorites. Perhaps weeding an online link collection shares more with an academic library weeding approach, but having never worked in that environment, I'll open the floor to the wisdom of others.