Prediction of Unsolved Terrorist Attacks Using Group Detection Algorithms

Authors:
Fatih Ozgul;Zeki Erdem;Chris Bowerman
Affiliations:
Department of Computing &Technology, University of Sunderland, Sunderland, United Kingdom SR6 0DD;Information Technologies Institute, TUBITAK- Marmara Research Centre, Gebze, Turkey 41470;Department of Computing &Technology, University of Sunderland, Sunderland, United Kingdom SR6 0DD
Venue:
PAISI '09 Proceedings of the Pacific Asia Workshop on Intelligence and Security Informatics
Year:
2009

Citing 4
Cited 0

Data Mining: Concepts, Models, Methods and Algorithms

Data Mining: Concepts, Models, Methods and Algorithms
Graph-based technologies for intelligence analysis

Communications of the ACM - Homeland security
Mining for offender group detection and story of a police operation

AusDM '07 Proceedings of the sixth Australasian conference on Data mining and analytics - Volume 70
Comparing Two Models for Terrorist Group Detection: GDM or OGDM?

PAISI, PACCF and SOCO '08 Proceedings of the IEEE ISI 2008 PAISI, PACCF, and SOCO international workshops on Intelligence and Security Informatics

Quantified Score

Hi-index	0.00

Visualization

Abstract

Detection of terrorist groups using crime data has few examples currently; this is because of lack of detailed crime data which contain terrorist groups' attacks and activities. In this study, a novel prediction model; CPM is applied to a crime dataset which includes solved and unsolved terrorist events in Istanbul, Turkey between 2003 and 2005, aiming to predict perpetuators of terrorist events which are still remained unsolved. CPM initially learns similarities of crime incident attributes from all terrorist attacks and then puts them in appropriate clusters. After solved and unsolved attacks are gathered in the same "umbrella" clusters, CPM classifies unsolved cases to previously known terrorist groups. Recall and precision results and findings of CPM regarded as successful then a baseline system; TMODS.