Transforming Data Product Generation through Federated Learning: An Exploration of FL Applications in Data Ecosystems

The significant increase in data generation across various sectors has prompted the development of concepts such as Data Product and Data Economy (DE) to enhance organizational productivity. Concurrently, advancements in AI models have heightened data privacy concerns, particularly as typical AI model training methods often involve data collection and storage in centralized databases, which are exposed to misuse. In response, Federated Learning (FL) has emerged as a promising approach, enabling the collaborative training of AI models without the direct sharing of data. This paper examines the potential of FL in the initial stages of data generation and throughout the data product design process. It further explores how FL can facilitate the generation of data products, providing a range of practical applications across different industries to address privacy concerns effectively in modern AI solutions.