Posts Tagged 'semantic web'

“Info Systems Must ‘Connect Dots’ on Terrorism” by Mannes and Hendler

There’s a new article in Defense News making some of the same points I’ve been making here: that accumulating data is not enough; what’s missing is tools to process the data and point out to human analysts what and where the important stuff is.

Mannes and Hendler make two good points: that it’s not about more data; and it’s not about being able to fuse data from different sources, although these are both good things. The problem is that there is still too much data to find out anything interesting, especially in a timely way.

I do believe that it’s important and useful to be able to automatically deduce what kind of object a given string describes, and so hypothesise what its attributes might be — so that people names come with locations and relatives. I’m not convinced that this has much to do with the Semantic Web (and I’m fairly sure you can generate a knock down fight by asking any two Semantic Web researchers to give a precise definition of what it is).

It’s sad, in a way, that the technology that is used as the ubiquitous metaphor for post-9/11 knowledge discovery (connecting the dots) is still the piece that isn’t being done! (I made the same point about the Australian government’s white paper which has a gaping hole about exactly this issue.)