Supporting Requirements Engineers in Recognising Security Issues

摘要

Context & motivation: More and more software projects today are security-related in one way or the other. Many environments are initially not considered security-related and no security experts are assigned. Requirements engineers often fail to recognise indicators for security problems. Question/problem: Ignoring security issues early in a project is a major source of recurring security problems in practice. Identifying security-relevant requirements is labour-intensive and error-prone. Security may be neglected in order to finish on time and in budget. Principal ideas/results: In this paper, we address this problem by presenting a tool-supported method that provides assistance for requirements engineering, with an emphasis on security requirements. We investigate whether security-relevant requirements can be automatically identified using a Bayesian classifier. Our results indicate that this is feasible, in particular if the classifier is trained with domain specific data and documents from previous projects. Contribution: We show how the ability to identify security-relevant requirements can be integrated in a workflow of requirements analysis and reuse of experience. In practice, this can increase security awareness within the software development process. WE- discuss limitations and potential of this approach.