Cancer models used to mine Facebook

As all who read Downtime know, technology tends to be pretty predictable (Moore’s Law still holds), but what people will do with it remains a mystery.

Take the latest modelling tool from the University of Warwick’s department of statistics and centre for complexity science (aka the chaos class). They invented this cool bit of technology called “graphical models” to unpick the complex cell interactions that lead to cancer.

No sooner had they published some results that some new kids on the block, aka sociologists, purloined it to datamine Facebook.

The boffins at Warwick looked at just 14 proteins which were implicated in the development of a form of cancer. Those 14 proteins could get together in around 8.72 followed by 10 zeros, according to Downtime’s rusty HP12C.

So, to save MIPS, they found a way using “informative priors” to interpret new information in light of what is already known. The clever bit is that the system will even recover from earlier errors in understanding.

Not only was the network model with information priors much better able to resolve complex interactions, it was much more accurate, they said.

They are now looking at how the tool could be used to mine anonymised data from social networking sites to better understand large scale interactions and relationships in society at large. Downtime expects the political parties have already placed their orders.