CrimeNet Explorer: A Framework for Criminal Network Knowledge Discovery

CrimeNet Explorer: A Framework for Criminal Network Knowledge Discovery
Key: central member, subgroups, interaction and relationship patterns between groups
Key method: structural analysis

The framework included four stages: network creation, network partition, structural analysis, and network visualization&&network visualization

Structural analysis methods in SNA fall into two categories: relational and positional
1.relational analysis
Relational analysis focuses on relationships and interactions between network members. It is often used to identify central members or to partition a network into subgroups.

three most popular centrality measures: degree, betweenness, and closeness.
(A node's score of beteweenness and closeness required computing the shortest paths(Dijkstra))

2.positional analysis
Purpose: discover the overall structure of a social network

Key approach :  blockmodeling (with hierarchical clustering ).
(it could be used to identify interaction patterns between subgroups discovered in the network partition stage)
includes two steps: network partition and interaction pattern identification.

key words: structural equivalence measure,complete-link algorithm

hierarchical clustering:the complete-link algorithm is the most popular

Visualization of Social Networks:
use MDS to visualize networks( the key step is SVD which could be solved using the library routine provided by Press et al.)

It seems that different strategies should be used to disrupt the different types of networks.(Based on the examples of gang networks and narcotics networks)

For testing a system: precision, recall, effectiveness, efficiency should be taken into cinsideration