Towards a federated learning approach for privacy-aware analysis of semantically enriched mobility data

Dagdia Z. C., Renso C., Zeitouni K., Agoulmine N.

Privacy  [INFO.INFO-NI]Computer Science [cs]/Networking and Internet Architecture [cs.NI]  [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG]  Mobility  Federated Learning  Trajectories  Mobility data 

Today, Artificial Intelligence is still facing a major challenge which is the fact of handling and strengthening data privacy. This challenge rises from the collected data which are associated with the fast development of mobile technologies, the huge capacities of high performance computing, and the large-scale storage in the cloud. In this paper, we focus on a possible solution to this challenge which is the use and application of federated learning. Specifically, beyond the federated learning based approaches proposed in different application domains, we mainly focus and discuss a federated learning approach for privacy-aware analysis of semantically enriched mobility data. We introduce the main motivation and opportunities of applying federated learning in mobility data, and highlight the main concepts and basics of our approach by describing our objectives and our approaches' requirements. We, also, describe our workplan that will permit achieving our predefined objectives via the setup of several research questions.

Source: FRAME '21: 1st Workshop on Flexible Resource and Application Management on the Edge, pp. 17–20, Online Conference, 25/06/2021

Metrics