Research Article

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

Table 6

Sensitivity analysis results under the characteristic node of Bayesian network.

Bayesian nodesState of Bayesian nodesEstimated probability
B 1 D 1B 1 D 2B 1 D 3B 1 D 4B 2 D 1B 2 D 2B 2 D 3B 2 D 4

(G) Cheat typeG 10.120.140.150.110.200.070.130.17
(G) Cheat typeG 20.120.110.150.240.170.250.150.30
(G) Cheat typeG 30.270.320.170.160.220.090.160.09
(G) Cheat typeG 40.170.100.140.150.160.060.150.20
(G) Cheat typeG 50.110.060.130.160.110.150.140.12
(G) Cheat typeG 60.150.230.140.150.100.330.140.08
(G) Cheat typeG 70.050.030.130.030.040.060.130.04
(I) Call the policeI 10.960.960.960.960.960.960.960.96
(I) Call the policeI 20.040.040.040.040.040.040.040.04
(J) Property lossJ 10.470.450.480.460.480.440.470.48
(J) Property lossJ 20.160.160.160.160.160.160.160.18
(J) Property lossJ 30.220.220.200.220.200.240.210.20
(J) Property lossJ 40.070.070.060.070.060.060.060.06
(J) Property lossJ 50.090.100.100.090.090.110.100.09