Risk Modeling for Terrorist Attacks

Hey guys, this will be the weekly post for the Earthquakes group. We've talked about a lot of different types of disasters so far, and this week we're focusing on terrorist attacks as disasters. I came across an article about a recent incident in Kunming, China, that has been labeled by the government as a terrorist act. A group of eight people armed with knives attacked a railway station, killing at least 28 people and wounding 113. The attackers killed at random and gave no warning prior to the attack. As of now there is still a lot of confusion surrounding the attack, but it is believed to have been the work of a separatist group in Xanjiang, in the northwest of China. The Uighurs are an ethnic minority in Xinjiang whose members, primarily Turkic-speaking Muslims, are engaged in an ever-growing conflict with China’s ethnic Han majority. The Uighers complain of discrimination from the Han, and many want to create their own separate state called East Turkestan in northwest China. Terrorist attacks are much different from natural disasters because there are very few ways of predicting or preventing them. In creating a risk model to help predict when and where a natural disaster will strike, we can better prepare for its aftermath. However, it is harder to predict human behavior than the weather. In his article "Risk Modeling Advances for Natural Disasters, Terrorism," Ben Dipietro argues that in order to create a model for terrorist attacks, "researchers need to try to understand the motivations of someone pursuing an attack, the logistics available to someone to launch an attack as well as what would be a high-profile target." 
As we already know, terrorist attacks are focused on civilians, and their effect is mainly psychological. Terrorists usually try to use these attacks as leverage against governments to try and attain larger political goals. If we were to try and make a risk model for terrorism, we would first have to understand the motivations of the people planning an attack, including both their short-term and long-term objectives. For example, while the Uighers want to prove to the Chinese government that they should be taken seriously, and to the Han people that they are a legitimate threat, but they also want sovereignty and statehood in the long-term. They are more likely to attack a large city in a public place where they will be guaranteed media coverage. Although weapons such as bombs may cause more casualties, these separatists chose to attack with knives--why? Looking back, there have been multiple mass knife attacks in China in the past three years, and inciting terror, as those attacks did, may have been a more important motivational factor to these separatists rather than killing the most people.
In the aftermath of the attack, it is relatively easy to see what caused it, but hindsight is 20/20. The question remains: how accurately can we predict the actions of terrorists? To what degree can governments prevent these types of attacks? What are some types of models we can use or create that would aid these types of disasters?

Here is the CNN article: http://www.cnn.com/2014/03/02/world/asia/china-railway-attack/
Here is the risk modeling article: http://blogs.wsj.com/riskandcompliance/2014/03/04/risk-modeling-advances-for-natural-disasters-terrorism/