A Victim-Based Framework for Telecom Fraud Analysis: A Bayesian Network Model

Table 2

Design of Bayesian network nodes for telecom fraud.

Nodes type	Nodes (BN variables)	States of Bayesian nodes
Portrait	(A) Sex	(1) Male, (2) female
	(B) Age	(1) Youth, (2) middle age, (3) old age
	(C) Marriage	(1) Married (2) unmarried
	(D) Work	(1) Company, (2) school, (3) selfemployed person, (4) government
	(E) Knowledge	(1) High, (2) low
Fraud process	(F) Cheat type	(1) Identity fraud, (2) shopping fraud, (3) inducement fraud, (4) fictional dangerous situation fraud, (5) daily consumption fraud, (6) phishing and Trojan virus fraud, (7) other types of cheat
	(G) Community type	(1) Phone, (2) message, (3) social software
	(H) Suspect during cheat	(1) Yes, (2) no
	(I) Call the police	(1) Yes, (2) no
Scam results	(J) Property loss	(1) 0–1000; (2) 1000–5000; (3) 5000–20000; (4) 20000–50000; (5) 50000+