An Approach to Predict SOA Security Vulnerabilities using Feed Forward Artificial Neural Networks

Abstract

The Service Oriented Architecture (SOA) makes interconnection of related applications in an organization easy and the web services provide platform and language independent implementation of SOA as they are using the basic Internet protocols for data transmission. As the data is transferred between servers and clients in a plain text format, it adds more security vulnerabilities to the existing security threats. It is increasing due to the latest web development such as Web 2.0 and its upgraded version Web 3.0; the traditional transport layer security solutions are totally absolute. Also even if specific solutions will be developed for the known security attacks, it is not possible to create a comprehensive one as the security risks are increasing and the problem space is wide. In this paper, we propose an approach to predict SOA security vulnerabilities in an organization using Artificial Neural Networks by applying customized learning techniques.