Machine learning evaluation of the requirement engineering process models for cloud computing and security issues

Abstract

In the requirement engineering phase, the team members work to get the user requirements, comprehend them and specify them for the next process. There are many models for the requirement engineering phase. There is a need to select the best Requirement Engineering model, and integrate it with cloud computing, that can give the best response to the users and software developers and avoid mistakes in the requirement engineering phase. In this study, these models are integrated with the cloud computing domain, and we report on the security considerations of all the selected models. Four requirement engineering process models are selected for this study: the Linear approach, the Macaulay Linear approach, and the Iterative and Spiral models. The focus of this study is to check the security aspects being introduced by the cloud platform and assess the feasibility of these models for the popular cloud environment SaaS. For the classification of the security aspects that affect the performance of these model, a framework is proposed, and we check the results regarding selected security parameters and RE models. By classifying the selected RE models for security aspects based on deep learning techniques, we determine that the Loucopoulos and Karakostas iterative requirements engineering process model performs better than all the other models.