Federated Continual Learning for Text Classification via Selective Inter-client Transfer
This work addresses the challenge of improving deep learning models over time for clients in cloud-edge environments without sharing data, representing an incremental advancement by applying federated continual learning to NLP for the first time.
The paper tackles the problem of combining federated learning and continual learning for text classification by proposing a framework that selectively transfers knowledge between clients to minimize interference from heterogeneous tasks, resulting in an average performance gain of 12.4% over a sequence of tasks using five diverse datasets.
In this work, we combine the two paradigms: Federated Learning (FL) and Continual Learning (CL) for text classification task in cloud-edge continuum. The objective of Federated Continual Learning (FCL) is to improve deep learning models over life time at each client by (relevant and efficient) knowledge transfer without sharing data. Here, we address challenges in minimizing inter-client interference while knowledge sharing due to heterogeneous tasks across clients in FCL setup. In doing so, we propose a novel framework, Federated Selective Inter-client Transfer (FedSeIT) which selectively combines model parameters of foreign clients. To further maximize knowledge transfer, we assess domain overlap and select informative tasks from the sequence of historical tasks at each foreign client while preserving privacy. Evaluating against the baselines, we show improved performance, a gain of (average) 12.4\% in text classification over a sequence of tasks using five datasets from diverse domains. To the best of our knowledge, this is the first work that applies FCL to NLP.