Research on Applying the Theory of Wavelet Neural Network and Entropy-Grey Association to the Security Risk Assessment for E-Government Information System

Abstract

The theories and methods of security risk assessment for e-government information systems have become one of the study focuses in the field. As the traditional security risk assessment method is more subjective in the aspect of assigning weights, the assessment results are often inconsistent with the reality. In order to overcome the problem, this paper presents a risk assessment method which combines wavelet neural network (WNN) and entropy-grey correlation, creates a WNN model and describes a simulation experiment by Matlab 7. In addition, comparisons are made in terms of convergent speed, training precision and forecasting effect between WNN and other traditional estimation methods such as BP-NN (Back Propagation Neural Network), FCM (Fuzzy Clustering Method) and SPR (Statistical Pattern Recognition). The experiment results indicate that, with the proposed method, subjectivity in the previous research is effectively avoided and the evaluation results are more scientific and valid. Case study proves that the method can be easily applied to the security risk assessment of the e-government information systems to produce objective results.