A Computationally-Enabled Analysis of Lashkar-e-Taiba Attacks in Jammu and Kashmir

Lashkar-e-Taiba (LeT for short) is one of the deadliest terrorist groups in the world. With over 100 attacks worldwide since 2004, LeT has become a political force within Pakistan, a proxy militia for the Pakistani Army, and a terror group that can carry out complex, coordinated attacks such as the 2008 Mumbai attacks. We have collected 25 years of data about LeT starting in 1985 and ending in 2010. The data is recorded on a monthly basis and includes the values of approximately 770 variables for each month. The variables fall into two categories -- action variables describing actions taken by LeT during a given month and environmental variables describing the state of the environment in which LeT was functioning. Based on this data, we have used our Stochastic Opponent Modelling Agent (SOMA) platform to automatically learn models of LeT's behavior. These models describe conditions under which LeT took various actions -- more importantly, the conditions act as predictors of when they will take similar actions in the future. In this paper, we focus on attacks by LeT in Jammu & Kashmir. We describe some conditions under which LeT ramps up offensive activities in Jammu & Kashmir. We conclude with some policy options that may reduce the use of violence by LeT as indicated by the rules presented here.

[1]  Ryan Clarke,et al.  Lashkar-I-Taiba: The Fallacy of Subservient Proxies and the Future of Islamist Terrorism in India , 2010 .

[2]  Philip A. Schrodt.  Forecasting Conflict in the Balkans using Hidden Markov Models , 2006 .

[3]  V. S. Subrahmanian,et al.  The SOMA Terror Organization Portal (STOP): social network and analytic tools for the real-time analysis of terror groups , 2008 .

[4]  V. S. Subrahmanian,et al.  CAPE: Automatically Predicting Changes in Group Behavior , 2009 .

[5]  V. S. Subrahmanian,et al.  A stochastic language for modelling opponent agents , 2006, AAMAS '06.

[6]  V. S. Subrahmanian,et al.  Stochastic Opponent Modeling Agents : A Case Study with Hamas , 2008 .

[7]  J. Stern,et al.  Terror in the Name of God: Why Religious Militants Kill , 2003 .

[8]  Samir Khuller,et al.  Finding Most Probable Worlds of Probabilistic Logic Programs , 2007, SUM.

[9]  A. Clauset,et al.  On the Frequency of Severe Terrorist Events , 2006, physics/0606007.

[10]  V. S. Subrahmanian.  Cultural Modeling in Real Time , 2007, Science.

[11]  Jeffrey H. Norwitz,et al.  The Counter-Terrorism Puzzle: A Guide for Decision Makers , 2006 .

[12]  Samir Khuller,et al.  Computing most probable worlds of action probabilistic logic programs: scalable estimation for 1030,000 worlds , 2007, Annals of Mathematics and Artificial Intelligence.

[13]  Diego Reforgiato Recupero,et al.  CARA: A Cultural-Reasoning Architecture , 2007, IEEE Intelligent Systems.

[14]  V. S. Subrahmanian,et al.  Stochastic Opponent Modeling Agents: A Case Study with Hezbollah , 2008 .

[15]  Wilson John,et al.  Investigating the Mumbai Conspiracy , 2009 .

[16]  F. W. Wiegel,et al.  A Generalized Aggregation-Disintegration Model for the Frequency of Severe Terrorist Attacks , 2009, 0902.0724.