Want to predict terrorist behavior?
The prediction of terrorist behavior is the goal of a new initiative at the University of Maryland. The University launched a data mining portal for counter-terrorism research.
The researchers consider that the unpredictable groups are useful for policy analysts and counter-terrorism groups who can use past behavior to forecast terrorist behavior.
The University of Maryland's Institute for Advanced Computer Studies (UMIACS), specifically, the SOMA Terror Organization Portal (STOP), uses publicly available data on more than 110 terror groups from around the world. In addition, it uses a real-time data extraction tool called T-REX to scour and extract data from more than 128,000 articles a day on an average of 180 news sites in 93 countries. The data is then organized into columns by year, variables associated with the group, such as an attack it might have carried out, or any counter-measures taken against it by a government. As a result, each variable then gets a numeric code representing its relative importance.
SOMA, or Stochastic Opponent Modeling Agents, then creates rules about the various terrorist groups, thus predicting their behavior, in its database.
The conclusions are worth considering. Hezbollah is demonstrated then to show that when it was involved in electoral politics, the chances it will attack civilians outside of Lebanon was in the 69% to 87% range. On the other hand, those chances dropped sharply when Hezbollah is not involved in electoral politics. The conclusion seems to be counterintuitive. Hezbollah is more violent as it is involved in democratic politics.
SOMA proved to be accurate in predicting an outcome about 90% of the time. This accuracy rate would be invaluable if consistent. The researchers had inputted ten years of data on each group and as a result turned out an accuracy rate over 90%. While the tool could not predict any specific target or time line the data could be invaluable for increasing security.
Nonetheless, the tool is a promising beginning for generally baffling human phenomenon.