Human-centered design based on federated learning to optimize user experience on blockchain platforms
In modern decentralized environments, designing systems that are not only efficient and secure but also understandable and trustworthy for end-users has emerged as the central challenge. Blockchain and Federated Learning (FL) have become two seminal technologies that facilitate privacy-preserving and transparent processing of data over distributed networks. Nevertheless, although most current Blockchain-FL architectures possess technical strengths, they are still complex, opaque, and challenging for end-users to engage with, causing low trust and limited usage. In order to fill the gap, this study discusses a Human-Centered Design (HCD)–based framework that couples blockchain and federated learning to reconcile technological robustness with human usability. The framework starts by exploring user needs and designing usable, translucent interfaces, then distributing model training without raw data sharing, while blockchain provides transparency and immutability of updates. User feedback gathered by standard User Experience (UX) metrics—System Usability Scale (SUS) and User Experience Questionnaire (UEQ)—instruct iterative interface and training workflow refinements. Experimental assessments performed over three real-world healthcare datasets—Heart Disease, Breast Cancer Wisconsin, and COVID-19 CT—show that the proposed framework enhances model accuracy by 4.1%, reduces latency by 12.5%, decreases communication cost by 24.6%, and improves the fairness index by 9.7% compared with the FL-only and FL + Blockchain without HCD baselines. In addition, the UX evaluation achieved an SUS score of 89 and a UEQ value of 6.3, indicating a 23% improvement in user satisfaction and trust. In totality, the result verifies that incorporating HCD wisdom into Blockchain-FL systems closes the gap between technical robustness and human experience, preparing the path toward transparent, secure, and user-friendly decentralized systems.