Back to Search View Original Cite This Article

Abstract

<jats:p>The data-driven and black-box nature of artificial intelligence (AI) systems have made the specification and evaluation of their requirements more intricate than other systems. A neglected topic is adopting a holistic approach to the quality requirements of AI systems. These systems also can lack explainability due to their architecture and the complexity of learning models. Therefore, modeling the quality requirements of AI systems can offer various benefits for successful design and development. In a case study, the internal quality requirements of an AI-integrated e-commerce system are modeled according to the Systems and Software Quality Requirements and Evaluation standard. The maintainability quality characteristic and its modularity sub-characteristic formed the quality focus. Systems Modeling Language offered rich semantics for capturing different aspects of the quality requirements. It is hoped that this holistic and integrated quality approach will enhance the construction of AI systems by ensuring high-quality specifications and more accurate and trustworthy AI solutions.</jats:p>

Show More

Keywords

systems quality requirements evaluation their

Related Articles

PORE

About

Connect