Improving Usability of Explainable Artificial Intelligence (XAI) Applications Through Quality Assessment Using Quality Model

The evolution of explainable artificial intelligence (XAI) software, which emphasizes seamless interaction with users to support their daily activities, presents a fresh challenge in evaluating system quality. Distinctive features such as invisibility and trustworthiness render it difficult to align with a generic quality model for assessing its effectiveness. Consequently, this endeavor seeks to identify the quality characteristics and model necessary for evaluating XAI software. To achieve this objective, a systematic literature review will be conducted to pinpoint the quality traits of XAI software and existing metrics for assessing their usability. Subsequently, a usability measurement model will be devised and subjected to validation by experts through review sessions. The outcome will be an enhanced usability model tailored for XAI applications, serving as a reference for researchers when evaluating such systems using a specifically designed model.