Adding Value to the Blog Award Nomination Collections…

Why, oh why, don’t publishers of blog award nomination lists see them as potentially useful collections on a articular subject that can be put to work for the benefit of that community?

For example, from a post on Brian Kelly’s blog, I notice that it’s that time of year when the blog awards process fires up again, in this case, the Computer Weekly IT blog awards.

There are umpteen categories – each category has it’s own web page – and umpteen nominations per award. To my mind, lists of nominations for an award are lists of items on a related topic. Where the items relate to blogs, presumably with an RSS feed associated with each, the lists should be published as an OPML file, so you can at-a-click subscribe to all the blogs on a list in a reader such as Google Reader, or via a dashboard such as netvibes. Where there are multiple awards, I’d provide an OPML file for each award, and a meta-bundle that collects nominations for all the awards together in a single OPML file, though with each category in its own nested outline element.

So has Computer Weekly made OPML feeds available for each nomination? Not that I can see… And a quick look at the award nomination pages suggests that a simple scraper routine won’t necessarily work to pull out all the nominations from the separate pages… (I gave myself 20 mins to do create the OPML bundles (it shouldn’t take any longer…), but stopped when I realised I wouldn’t complete the task in that time because of the crap way the info was published;-)

As well as creating the OPML bundles to make it easy for folk to subscribe to the nominees by category (or just preview the sort of thing they cover via a Netvibes style dashboard), it’d be trivial enough to also create a Google Custom search engine to search over just the nominated blogs. Taking this approach – trying to turn blog award nomination lists into useful resources, and extract value from the collection – sees the collection as a valuable resource, and starts to explore ways of (making it easier to) extract value from those collections; (if it’s one thing I think the librarians have traditionally understood, it’s the idea of a good collection, even if they’re not so hot at working out what that means on the web;-)

By looking for value in the collection – e.g. the extent to which the nominated blogs turn up useful results in a search engine on “software development” or “IT Security” (both CW IT blog award areas) – we can see whether or not the nominated blogs are actually “valuable” and worth rewarding, at least in the usefulness stakes;-)

PS In addition to the blog awards, there’s an IT Twitterers award. I’m not even sure whether they’ve collected together a list of the nominated Twitterers as a Twitter list…? So out of the generosity of my heart, I wrote a little script for my newt library to create a twitter list from a text file containing a list of Twitter usernames:

def txtFileToList(api,o,tag,fname):
  for i in f:

and created a Computer Weekly IT Twitter User of the Year Award (Nominations) List. (Note that a couple of nominations on the list appear as typos… it’s left as an exercise for the reader to identify which ones;-)

Just because, here’s a look at how incestuous – or not – that list is, by virtue of the extent to which the nominees follow each other on Twitter:

How Computer Weekly nominated IT Twitterer of Year award nominees follow each other

(Node size is proportional to how many other nominees follow a user, colour/heat the number of nominees they follow.)

The modularity detection algorithm doesn’t identify many clusters – the network is maybe too tightly interconnected? – although it does identify a small group interested in security:

CW twitterers - clustering

(I was hoping the journalists might also fall out as a group, but the don’t seem to have..:-(

Anyway, enough time wasted on that for a morning, I think…