Sociocultural predictive analytics and terrorism deterrence

Antonio Sanfilippo, Liam McGrath

Research output: Contribution to journalArticle

Abstract

Frame analysis, which focuses on the communicative and mental processes, is used to develop a strategy for terrorism deterrence. The occurrence, or lack of frame annotations are used that are highly correlated with terrorist documents to train computational models that identify messages from terrorist sources and estimate their proximity to an attack. Communication differences between al Qa'ida (AQ) and Al Qaida in the Arabian Peninsula (AQAP) using various decision-tree (DT) algorithms with frame annotations from documents issued by AQ and AQAP sources were modeled. A strong correlation between group-member authorship and AQAP messages, and between group-leader authorship and AQ messages, is found. The alternating DT model shows that AQAP group leaders are more likely to target young Muslims, whereas AQ group leaders focus on other audiences.

Original languageEnglish
Article number5968109
Pages (from-to)87-91
Number of pages5
JournalIEEE Intelligent Systems
Volume26
Issue number4
DOIs
Publication statusPublished - Jul 2011
Externally publishedYes

Fingerprint

Terrorism
Decision trees
Communication
Predictive analytics

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Networks and Communications

Cite this

Sociocultural predictive analytics and terrorism deterrence. / Sanfilippo, Antonio; McGrath, Liam.

In: IEEE Intelligent Systems, Vol. 26, No. 4, 5968109, 07.2011, p. 87-91.

Research output: Contribution to journalArticle

@article{af849f7d44bc4f2b80d90f16228c6b9a,
title = "Sociocultural predictive analytics and terrorism deterrence",
abstract = "Frame analysis, which focuses on the communicative and mental processes, is used to develop a strategy for terrorism deterrence. The occurrence, or lack of frame annotations are used that are highly correlated with terrorist documents to train computational models that identify messages from terrorist sources and estimate their proximity to an attack. Communication differences between al Qa'ida (AQ) and Al Qaida in the Arabian Peninsula (AQAP) using various decision-tree (DT) algorithms with frame annotations from documents issued by AQ and AQAP sources were modeled. A strong correlation between group-member authorship and AQAP messages, and between group-leader authorship and AQ messages, is found. The alternating DT model shows that AQAP group leaders are more likely to target young Muslims, whereas AQ group leaders focus on other audiences.",
author = "Antonio Sanfilippo and Liam McGrath",
year = "2011",
month = "7",
doi = "10.1109/MIS.2011.74",
language = "English",
volume = "26",
pages = "87--91",
journal = "IEEE Intelligent Systems and Their Applications",
issn = "1541-1672",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "4",

}

TY - JOUR

T1 - Sociocultural predictive analytics and terrorism deterrence

AU - Sanfilippo, Antonio

AU - McGrath, Liam

PY - 2011/7

Y1 - 2011/7

N2 - Frame analysis, which focuses on the communicative and mental processes, is used to develop a strategy for terrorism deterrence. The occurrence, or lack of frame annotations are used that are highly correlated with terrorist documents to train computational models that identify messages from terrorist sources and estimate their proximity to an attack. Communication differences between al Qa'ida (AQ) and Al Qaida in the Arabian Peninsula (AQAP) using various decision-tree (DT) algorithms with frame annotations from documents issued by AQ and AQAP sources were modeled. A strong correlation between group-member authorship and AQAP messages, and between group-leader authorship and AQ messages, is found. The alternating DT model shows that AQAP group leaders are more likely to target young Muslims, whereas AQ group leaders focus on other audiences.

AB - Frame analysis, which focuses on the communicative and mental processes, is used to develop a strategy for terrorism deterrence. The occurrence, or lack of frame annotations are used that are highly correlated with terrorist documents to train computational models that identify messages from terrorist sources and estimate their proximity to an attack. Communication differences between al Qa'ida (AQ) and Al Qaida in the Arabian Peninsula (AQAP) using various decision-tree (DT) algorithms with frame annotations from documents issued by AQ and AQAP sources were modeled. A strong correlation between group-member authorship and AQAP messages, and between group-leader authorship and AQ messages, is found. The alternating DT model shows that AQAP group leaders are more likely to target young Muslims, whereas AQ group leaders focus on other audiences.

UR - http://www.scopus.com/inward/record.url?scp=80051537314&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80051537314&partnerID=8YFLogxK

U2 - 10.1109/MIS.2011.74

DO - 10.1109/MIS.2011.74

M3 - Article

AN - SCOPUS:80051537314

VL - 26

SP - 87

EP - 91

JO - IEEE Intelligent Systems and Their Applications

JF - IEEE Intelligent Systems and Their Applications

SN - 1541-1672

IS - 4

M1 - 5968109

ER -