BIG DATA ANALYTICS BASED ON LINKED LEARNING FOR SMART CITY APPLICATIONS THAT PROTECT PRIVACY

In the context of smart city systems generating large-scale and distributed data, ensuring efficient data analytics while preserving privacy has become a critical challenge. This paper proposes a big data analytics framework based on Federated Learning (FL), enabling distributed model training without sharing raw data. The proposed architecture integrates IoT devices, edge computing, and a federated server, combined with an enhanced FedAvg algorithm to address data heterogeneity. In addition, privacy-preserving mechanisms such as Differential Privacy and Secure Aggregation are incorporated to strengthen data protection. Experimental results on IoT and smart city datasets demonstrate that the proposed model achieves performance comparable to centralized approaches while maintaining strong privacy guarantees. The approach shows high potential for practical deployment in distributed big data environments.