Wednesday, September 10, 2008


Had a look at this. It's a nice idea, a site that lets you create a search tool that works across your favourite sites, but in practice the results it brings back aren't that great, and on the whole seem to be a pretty good demonstration of why librarians are actually still needed when it comes to searching out targetted info using sites specifically developed for that purpose.

A metasearch tool like this is always going to struggle when it is being asked to operate across a disparate range of sites with little behind-the-scenes consistency. I probably won't use it - actually, I found it pretty irritating - as I wasn't that impressed with the search results it brought back from the Rollyo of local history sites I created. Why?
1. The only successful searches were with very generic searches that were too unfocussed to be useful - anything more specific and you got nothing.
2. It only searches text on the site itself, and if you work primarily with sites that are interfaces to behind-the-scenes databases, it isn't able to pick up info at the individual record level.

Can I see a role for it? No, not if you are wanting to search online databases, like our Local History Online, or Matapihi, or Auckland Museum's "Street Search" - but if all you want to do is pull together info actually written on the site itself in xml or whatever, then yes, it could be useful.

Not sure whether this link will work for other people, but anyway, here it is:

1 comment:

NSL Learning 2.0 team said...

Your link worked. Interesting ovservations in regards when Rollyo would be useful and when it won't work, which we need to keep in mind when looking at how Rollyo could be utilised in a reference setting.