From COSNet

Mind Games at Kioloa 2008 Home The Problem Set Working Groups Readings Presenters Mentors Participants Workshop Program Transport Things to Bring Accommodation Internet and Telephone Access General Safety Precautions ANU Kioloa Coastal Campus Wiki Primer Mock Media Release Reflections/Comments Photos

Contents 1 Press Release 2 Introduction 3 Research Questions for Kioloa 4 Meeting Notes 4.1 Wed AM 4.1.1 General Discussion 4.1.2 Way forward? 4.1.3 Information Needed 4.2 Thur AM 5 Data Set Analysis 5.1 Characterising the data-set 5.1.1 Columbia 5.2 Spatial Data 5.3 Event Type Transitions 6 Background 6.1 Data 6.1.1 Example 6.2 Working in the field 6.3 Reading

Press Release

VIRTUAL WORLDS TO MAKE INROADS ON REAL TERRORISM
------------------------------------------------

Computer games have long been a popular outlet for millions of people 
world-wide, but a group of young Australian scientists believe computer games 
can say a lot about the reality of global terrorism.

The newly formed Counter Terrorist Intelligence Unit (CTIU) -- a multi-university, 
cross-disciplinary team of physicists, mathematicians, economists, computer 
scientists and engineers -- represents the cutting edge in the new science of 
complexity.

The group has met for an intensive research thinktank at the Austrlaian 
National University's Kioloa Coastal Field Research facility, on the south 
coast of NSW.

They hope to build the first counter terrorist virtual world, where 
terrorists, civilians and counter-terrorist agencies, interact and vie for 
supremecy.

In a first step, they have been scouring a huge database of over 62,000 
terrorist events which occured around the world over the last 
25 years.  "Our first challenge was really to understand the patterns of 
global terrorism -- where terrorists come from, what types of attacks they 
carry out and of course, what kind of damage they inflicted," said Senior 
Research scientist, Kate Simms.

Their efforts have resulted in some promising early results which they hope 
will form the basis of their virtual counter-terrorist world. Like the massive 
multi-user computer games of 'Second Life' and 'There', the researchers 
believe that only by allowing virtual individuals to inhabit virtual 
landscapes will the complex relationships between terrorists and 
counter-terrorists be truly understood.

"Counter-terrorism has been for too long considered a responsive activity but 
this approach has been shown to be largely ineffective. By building a viritual 
world where terrorists, counter-terrorist authorities and civilians live, we 
can really start to understand how civilians progress from initial recruitment 
to final deployment," explained CTIU Team Leader, Holly Sims.

The group has received special fast-track funding from the Federal Government 
and will report on their work early in 2009.

Media contact: Name, Phone

Introduction

• What is the complex systems nature of global terrorism?
• Can it be defeated by a complex systems approach?
• Or can it only be coped with?

Research Questions for Kioloa

Meeting Notes

Wed AM

General Discussion

Why does a person become a terrorist?

• Many factors (dimensions)
  • religion
  • economic considerations (e.g. poverty, lack of job opportunities)
  • political
  • historical (possibly passed from parent to child)
  • demographics (e.g. poor elder to youth ratio)
  • education (e.g. literacy)
• are there any rich terrorists?
  • rich countries -- such as Egypt, still have terrorists, why?
  • regional policies cause some areas to be very poor, despite the overall riches of the country
  • a political failure?
  • inequality
  • lack of rule of law

What could we focus on?

• How a person takes on a belief -- how it is transferred in the society
  • also role of propaganda
• Prediction of terrorism?
  • is it possible?
  • isn't it a purely cultural/religious phenomenon

Case study: Northern Ireland

• why did the terrorist activity reduce here?
  • wealthier people because of the booming Irish economy?
  • where have the terrorists gone?
• are the cultural/religious divisions still bubbling away under the surface?

Interaction Between Terrorism and Response

• can get a disproportionate response against a terrorist activity by the authorities
• this causes an increase in the awareness of the terrorist activity, and possible increase in radicalised citizens who suffer under the knee-jerk/oedr-kill response of the authorities

What is the basis of conflict?

• land, religion, women? .. conflict over 'assets'

Where do you put your efforts?

• if education of women is very important .. do you focus on (say) opening schools rather than centralised military action

Bar-Yam: Predicting Ethnic Conflict

• Tolerance to other ethnicities a good predictors
• Need different main groups pressed up against each other

Way forward?

1. Focus on causes of terrorism, radicalisation?
2. Network approaches -- how are terrorist groups organised?
   1. what is the best way to defeat such organisations? .. network brittleness etc.
3. Agent modelling -- Aggressors and Defenders .. co-evolutionary approach to this

Information Needed

1. What is the 'colour' of terrorism?
   1. use the data set to get some interesting characteristics.
2. Who are terrorists?
   1. why do they become terrorists?
3. What current academic research is there about agent-based modelling, complexity and terrorism?
   1. a literature search on the net

Thur AM

Simon

• What happened in Columbia during the early 90s? .. why the reduction in the number of events?

Asef

• Paper in Complexity
• In the first case, consider one country, with power of terrorists T, and power of counter-terrorists CT
• They then use two components in their dynamics of T equation, one being the fact that the gradient/reduction of terrorists will be proportional to the number of Terrorists (i.e. more terrorists will cause bigger response of CT); and second, that if the CT activities are too harsh, then there will actually be an increase in T since civilians will become radicalised
• Idea:
  • does this miss the fact that T could actually increase proportional to T since the more terr. there are, the greater their ability to recruit

General Discussion

• Do we see a change in the data pre- and post- the widespread introduction of the internet
• Should we use the criminal element in a society as the 'control group' .. i.e. testing something like:

Terrorism Rate
--------------- = a0 + a1 A + a2 B + ... 
  Crime Rate

Roots of Terrorism

1. Terrorism -- poverty, is positive, or negative correlation?
   1. In some cases, can have very well educated, rich people who are terrorists
   2. Yet in (say) Saudi Arabia, some school provide students with everything -- food, books etc. .. and so training is from very early age to be against western interests
   3. Suicide bombers, often for economic reasons (i.e. from poor families)
2. Terrorism -- states .. failed, or good governance
   1. Movement of resources, people, drugs etc. between countries
   2. many terrorists come from poor countries and possibly become radicalised as a way to get to rich countries
3. Terrorism -- beliefs
   1. Many ME believe that T. is a very good/honourable things to do
4. Terrorism -- Finance
   1. State, or individual, or charity organisation (e.g. money from people to support T) will support the T. activities (esp. Saudi Arabia)
5. Terrorism -- Immigrants
   1. Immigrant from ME to Western country -- feel discriminated against and want to become T. to get back
   2. 3rd generation is often extreme generation .. not actually been to their homeland, don't really know their roots, and become strong advocates for their country of orgin
6. Terrorism -- Governance/Democracy
   1. Lack of democracy in a country can lead to the desire to become terrorist in these countries

Possible Research Approaches

1. Data-Econometric:
   1. Use the data-set to explore why/when terrorist events have occured -- explaining these via spatial, economic, demographic characteristics of the country
2. Agent-based, Simulation:
   1. Build a model of T and CT activities with a spatial element

Data Set Analysis

Image:Terrorism-code-v081120-1.zip

Characterising the data-set

Some points:

• Events largely accounted for by South America (Columbia, Peru, El Salvador), Western Europe (N Ireland, Spain)
• Peak in 1992 .. why?? .. decline in 94-97?
• This kind of data needs to be deflated by population changes.

Columbia

Some points:

• Most frequent action against Facilities (infrastructure?), Bombing and Assassination
• What happened around 1995? .. steep deep
  • Possible investigation of causes of this decline in attacks?
• Seems a good Power law fit in the WOUND data

Spatial Data

Simple Event map

Event Intensity map

Event Type Transitions

Cyclic nature of terrorism activities is an interesting stylized fact in the dataset

Colombia: Ireland:

Spain: ElSalvadore:

Background

Data

Data on terrorism events is available here:

• MIPT Knowledge base, alternative sources: ITERATE, GTD, ICT etc.
  • 32,829 events, spanning 40 yeras, 5600 cities, 187 countries
  • have to be known/corroborated by a number of sources
  • has to be politically motivated
  • 14,062 with casualties > 0

Or here:

• https://www.icpsr.umich.edu/ticketlogin
  • 61,637 events (rows) from 1970 -- 1997 and includes longitude and lattitude, plus whether the event was successful or not
  • Link to data (.csv, 13MB) (please use, this link will disappear soon)

Example

Working in the field

• Aaron Clauset (SFI)
• New England School of Complexity Science -- Ethnic conflict
• TTCP Action Group on Complex Adaptive Systems for Defence

Reading

(please add refs/links as you think-of/find them)

Terrorism and Complexity

• Media:Hayden2007-TerrorismComplexity.pdf

Directly related to terrorism

• On the Frequency of Severe Terrorist Attacks., A. Clauset, M. Young and K. S. Gledistch., Journal of Conflict Resolution 51(1), 58 - 88 (2007).
• Global Pattern Formation and Ethnic/Cultural Violence, M. Lim, R. Metlzer, and Y. Bar-Yam., Science 317, 5844 (September 14, 2007).
• Universal patterns underlying ongoing wars and terrorism. Neil F. Johnson, Mike Spagat, Jorge A. Restrepo, Oscar Becerra, Juan Camilo Bohorquez, Nicolas Suarez, Elvira Maria Restrepo, Roberto Zarama, arXiv, 2006.

Relevant science

• Power-law distributions in empirical data. A. Clauset, C. R. Shalizi and M. E. J. Newman. Submitted to SIAM Reviews (2007).

Complexity in war

• Characterising the Difference between Complex Adaptive and Conventional Combat Models. M. K. Lauren, arXiv, 2006.
• A Metamodel for Describing the Outcomes of the Mana Cellular Automaton Combat Model Based on Lauren's Attrition Equation. M. K. Lauren, arXiv, 2006.
• Land Warfare and Complexity, Part II: An Assessment of the Applicability of Nonlinear Dynamics and Complex Systems Theory to the Study of Land Warfare, A. Ilachinski, CNA Report, 1996.
• About the bears and the bees: Adaptive responses to asymmetric warfare. A. Ryan, Interjournal, 2006.
• Complexity Theory and Network Centric Warfare. J. Moffat, CCRP, Washington, USA, 2003.
• Clausewitz, Military Theory, and Nonlinearity
• Terrorism, Nonlinearity & Complex Adaptive Systems (note that many of these links are out of date)