A Controlled Natural Language Approach for Integrating Requirements and Model-Driven Engineering

Despite the efforts made during the last decades, Software Engineering still presents several issues concerning software products' quality. Requirements Engineering plays a important role regarding software quality, since it deals with the clear definition of the target system's scope. Moreover, Requirements Engineering is crucial to deal with change management, which is required to ensure that the final product reflects the stakeholders' expectations, namely the client and end-users business-related needs. We advocate the need to address the open issues regarding the requirements development process, namely to mitigate the drawbacks of using informal natural language, such as ambiguity and inconsistency. Moreover, we recognize the importance of automation to enhance productivity by avoiding repetitive and error-prone activities. In this paper, we propose a new socio-technical approach to overcome these software quality problems, consisting on the deep integration of Requirements Engineering with Model-Driven Engineering processes. This approach is based upon a controlled natural language for requirements specification, supporting the automatic extraction and verification of requirements models with Natural Language Processing techniques. The current results consist on the development of a Wild-based tool prototype to validate our research ideas.