An Ai-driven Human-centered User Experience Framework for Intelligent Systems: Design, Implementation, and Evaluation

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Artificial Intelligence (AI) has transformed intelligent systems by enabling adaptive, personalized, and automated user interactions across domains such as healthcare, education, finance, and e-commerce. However, many existing AI-powered interfaces prioritize algorithmic performance over usability, transparency, and user-centered design, limiting user trust and overall experience. This study proposes an AI-Driven Human-Centered User Experience Framework (AI-HUXF) that integrates Artificial Intelligence, Human–Computer Interaction (HCI), and User Experience (UX) principles to deliver adaptive, explainable, and personalized user interfaces. The framework incorporates user behavior analytics, machine learning-based personalization, adaptive interface generation, intelligent recommendations, and continuous learning to dynamically tailor interface components according to individual user preferences. A prototype was developed using React.js, Spring Boot, TensorFlow, and PostgreSQL and evaluated through controlled experiments using established usability metrics, including the System Usability Scale (SUS), user satisfaction, response time, task completion rate, error rate, cognitive load, and personalization accuracy. Experimental results demonstrate that the proposed framework shows higher reported performance than a conventional interface, achieving a SUS score of 93 compared with 78, increasing user satisfaction from 81% to 96%, reducing response time from 1.8 s to 0.9 s, improving task completion rate from 87% to 98%, and reducing the error rate from 8% to 2%. The AI personalization model further achieved 97.8% accuracy, 97.3% precision, 96.9% recall, 97.1% F1-score, and an AUC of 0.986, indicating high reported predictive performance for user-preference prediction. The findings demonstrate that AI-driven human-centered design shows improvements in the reported usability measures, personalization, and interaction efficiency, providing a scalable framework for developing trustworthy and intelligent next-generation user interfaces.