FLAWR 2024 : Federated Learning Applications in the Real World

Following a data protection by design principle, federated learning has emerged as a more privacy-friendly machine learning paradigm for independent actors to collaboratively train a machine learning model without sharing their local training data with a central server or other nodes. When applying federated learning in real-world scenarios, especially where there are complicated requirements on protection of local data (e.g., in health and crime-related data), a wide range of socio-technical aspects have to be considered together with technical ones. For instance, when federated learning is used to facilitate health data sharing between public hospitals and private companies, the system designers and implementers have to consider various aspects related to business models, legal compliance, regulatory processes, economic factors, human behaviours, and ethics, e.g., how relevant stakeholders especially patients and carers can be involved to ensure the transparency of the whole process and their actual implementations, how non-expert users perceive such technical solutions and adopt them, how complicated patient consents can be managed, how different subsets of data can be anonymised but remain linkable to allow more useful health analytics, and what ethical issues should be considered to better manage conflicting interests of different parties. Considering such socio-technical aspects in developing, deploying and evaluating federated learning to maintain real-world security and privacy requirements is not trivial, and often introduces complicated trade-offs that designers, developers and practitioners have to consider carefully.

This special session aims at providing a platform for researchers from different disciplines to share their latest research work on security and privacy aspects of practical applications of federated learning in the real world, with some interdisciplinary elements on one or more relevant socio-technical elements. We particularly welcome researchers from outside of Computer Science and Electronic Engineering to submit their work.

The list of possible topics includes, but is not limited to:
Security or privacy-critical applications of federated learning in different domains (e.g., health, energy, finance, policing, e-government)
Human factors in federated learning applications (e.g., usability, attitude towards adoption)
Behavioural aspects of real-world federated learning systems
Legal aspects of real-world federated learning systems
New business models enabled by federated learning
Economic aspects of federated learning (e.g., incentivisation, economic modelling)
Ethical considerations related to the use of federated learning
Real-world challenges of deploying federated learning and solutions
Societal impacts of federated learning in real-world applications
Socio-technical aspects of attacks on federated learning
Privacy risks or attacks of federated learning in real-world applications and solutions