Directed Terror Networks

It's such a truism to talk about terrorist networks these days. About the hubs and nodes and how the edges connect the vertices and three jumps take us right from Osama to Big Bird.

Yes, a lot of genius work is being done. Valdis Krebs is a master, and John Robb is my own personal Jesus. But there's also a lot of smack being talked.

I don't think it's enough to talk about terrorists as a network; I don't think it's useful to just dig deeper into those connections, either. For one thing, the amount of generated data is phenomenal, and even when you weight the relationships, there are still a lot of false leads.

So, I tried to step back and look at what it meant if we could define terrorist networks as specifically Directed Networks. Does framing the issue that way give us any new insight?

And yeah, I think it does.

So, a directed network is a network in which the connections aren't symmetrical. You click from one page to another and there's not immediately a route back. Like that.

In terms of social phenomena, I think it indicates that you can make relationship connections that it's hard to move back from.

Pair that with this idea: Most people are analysing these networks as if the nodes were actually people. But I think each node is actually the sum of the connections pouring into it. Yes, a person with a given set of qualities occupies it.

But remove that person and those connections are going to start looking for another node to latch onto. In a sense, then, people can move to occupy different nodes in a given relationship systems.

In a terrorist network, for instance, you might start off sympathizing, then supporting, then go operations. At each choide, you occupy nodes deeper and deeper into the network. After each choice, it's damn hard to get back where you were before.

This is really preliminary theory, rather than research. But to me, the idea of human networks and the concept of nodes as placeholder locations rather than people are both key to breaking down these nets.

Get the Paper

Comments

[red bottom]

So fun article is! I know more from it.

[Valdis]

I skimmed the paper -- looks VERY interesting. I will read in detail later this week.

I like your concept of changing roles or of "increased inclusion" into the network. Look at this recent network map I did:

http://orgnet.com/emergent_community2.png

This is an actual online community around a common affiliation here in the USA[these are NOT terrorists]. Nodes are colored by their involvment in the community.

The blue nodes are attracted to the community, but do not participate with anyone there -- just read the posted materials. They are the typical "lurkers". Or in your t-nets: sympathizers? Could this map of an emergent group also reveal something about how terrorists organize and attract sympathizers?

The green nodes are active locally within a small group [cells?]

The red nodes are the interconnected core of high activity [the al qaida base? ] We took most of al qaida's core out in Afghanistan. But in doing so, did the remaining nodes [that we did not get] become like the small green clusters in our map above? We took out the core but left the rest -- are we fighting a doughnut!!! ;-)

Again, the map you see[linked above] is an on-line community of positive folks here in the US. But their network shape could very well match how we see the bad guys, or any emergent human community.