Introduction to Algorithms
Evaluating similarity measures: a large-scale study in the orkut social network
Proceedings of the eleventh ACM SIGKDD international conference on Knowledge discovery in data mining
Use of ontologies and probabilistic relational models to aid in cyber crime investigation decision support
Knowledge Discovery for Counterterrorism and Law Enforcement
Knowledge Discovery for Counterterrorism and Law Enforcement
| Hi-index | 0.00 |
Some countries suffer from terrorism much more than others, Turkey as one of the most suffering countries who owns about a hundred terrorist groups; most of these organizations cooperate, and interchange knowledge, skills, materials used for terrorist attacks. From criminological perspective terrorist networks of Turkey are categorized into three main groups: extreme left (i.e. Marxist) networks, extreme right (i.e. Fundamentalist, Radical Islamist) networks, and separatist (i.e. ethic, racist) networks. By using their criminal history including the selection of crimes, attacking methods and modus operandi, a crime ontology is created, terrorist networks are attached to this ontology via their attacks and a similarity measure (COSM) is developed according to this ontology. Results of this similarity measure performed better than two common similarity measures; cosine and Jaccard. Results are also presented to domain experts in hierarchical clustering and they also commented as positive. Based on attributes of crimes, COSM similarity can also be applied to other types of social networks.