eurovision contest predictor based on search data

Last year, I ended up watching the Eurovision Song Contest because I wanted to take part in the conversations my friends were having about it on Twitter. In the resultant blog post, I looked forward to this year:

“Broadcasters and content providers should take note. What I
participated in last night would be almost totally invisible to most
viewers. Most people don’t know how to find and track conversations on
twitter, other social networking services or blogs. But being part of
an audience community is a powerful experience for participants and a
valuable brand building tool for broadcasters and other content
producers.

We need to make it as easy as possible for ordinary users to find
and participate in conversations around our content. The way to do that
isn’t to duplicate the tools and services that are already out there,
but to create interfaces, windows, that let people see and join into
the conversation. Underlying that interface there might be all sorts of
complex tools – hashtags, tweetscan, summize and twitterlocal are all
useful – but in pulling them all together in a meaningful way, much of
the complexity and need for prior knowledge is removed. Achieve that
and next year’s Europarty is going to be unforgettable.”

Well, it’s time for Eurovision again, but I’ve yet to see any good mash-ups or aggregations of conversations people are having around the contest or coverage of it but I have found an interesting piece of data mining by google, who has created a predictor gadget that looks at searches, filters in various ways for better accuracy, then ranks the contestants:

You might also want to follow tweets mentioning eurovision. Let me know if you find that mashup or aggregation I as hoping for.