Federated Incremental Named Entity Recognition

TL;DR

This paper introduces a federated incremental NER approach that effectively handles continuous new entity types and client dynamics, overcoming forgetting issues through a novel LGFD model with knowledge distillation and contrastive learning.

Contribution

The paper proposes a novel LGFD model for federated incremental NER, addressing intra- and inter-client forgetting with structural knowledge distillation, contrastive loss, and task switching monitoring.

Findings

LGFD significantly outperforms existing methods in incremental NER tasks.

The model effectively preserves old knowledge while learning new entity types.

Experimental results validate the robustness of LGFD in dynamic federated environments.

Abstract

Federated Named Entity Recognition (FNER) boosts model training within each local client by aggregating the model updates of decentralized local clients, without sharing their private data. However, existing FNER methods assume fixed entity types and local clients in advance, leading to their ineffectiveness in practical applications. In a more realistic scenario, local clients receive new entity types continuously, while new local clients collecting novel data may irregularly join the global FNER training. This challenging setup, referred to here as Federated Incremental NER, renders the global model suffering from heterogeneous forgetting of old entity types from both intra-client and inter-client perspectives. To overcome these challenges, we propose a Local-Global Forgetting Defense (LGFD) model. Specifically, to address intra-client forgetting, we develop a structural knowledge distillation loss to retain the latent space's feature structure and a pseudo-label-guided inter-type contrastive loss to enhance discriminative capability over different entity types, effectively preserving previously learned knowledge within local clients. To tackle inter-client forgetting, we propose a task switching monitor that can automatically identify new entity types under privacy protection and store the latest old global model for knowledge distillation and pseudo-labeling. Experiments demonstrate significant improvement of our LGFD model over comparison methods.