19 November 2004: Scholar

Quick note on something I should have registered yesterday, when Steve sent an email around to the Aktors group saying that Google were introducing Google Scholar, an academic papers portal. While relevant to our group, Google-style work tends to be significantly different from that of AKT in that we're interested in ontologies and they're not. But with Scholar, however, there's a particular twist that affects my side of this Akting lark.

Simply this: it's Citeseer. Albeit with less autonomous citation indexing (you get the feeling Google Scholar has a lot of hand-matched stuff), it almost certainly has (or soon will have) more content. Not to mention more access to the big journals and conferences. It seems that they'll have some content which is subscription-based: ie you can access the page headers and citations, but to read the actual article you'll need a subscription to the relevant body. So it's Citeseer-plus. As Isaac rightly said yesterday, it's certainly past time for a new Citeseer - just a shame we weren't the ones to get it done first.

Can't say this bodes well for Citeseer as a niche engine, which is a shame because it's a bleddy fine piece of work considering the small number of developers it's had over the years. From an AKT point of view it won't affect our work too heavily - it may even help it, depending on the nature of any web services or APIs Google choose to release. We're interested mainly in the citation graph and the metadata, so until Google release a version of Aktive Space, I think we'll be ok.

Still, not an encouraging day for the Citeseer community, small as we are.