A Review of Blockchain-Enabled Federated Learning Systems for Privacy-Preserving Artificial Intelligence

Recent advancements in intelligent computing techniques have significantly influenced the development of data-driven and automated systems across various application domains. As the volume and complexity of data continue to increase, selecting appropriate models and techniques has become a critical challenge for researchers and practitioners. This paper presents a comprehensive review of prominent intelligent and computational approaches reported in recent literature, focusing on their underlying principles, performance characteristics, and application suitability. The review systematically analyzes existing methods, highlighting their strengths, limitations, and comparative effectiveness based on key evaluation parameters. By examining multiple studies and frameworks, this work identifies common trends and recurring challenges associated with scalability, accuracy, and computational efficiency. Furthermore, the paper emphasizes existing research gaps and limitations that restrict the practical adoption of current approaches in real-world scenarios. The insights derived from this review aim to assist researchers in understanding the current state of the art and selecting suitable techniques for specific problem domains. Finally, potential future research directions are outlined to address the identified gaps and to encourage the development of more robust, efficient, and adaptive intelligent systems.