Mix of Experts Language Model for Named Entity Recognition
This work addresses noisy supervision in NER for natural language processing applications, representing an incremental improvement over existing distantly supervised methods.
The paper tackles the problem of incomplete and noisy annotations in distantly supervised Named Entity Recognition by proposing BOND-MoE, a model based on Mixture of Experts and Expectation-Maximization, which achieves state-of-the-art performance on real-world datasets.
Named Entity Recognition (NER) is an essential steppingstone in the field of natural language processing. Although promising performance has been achieved by various distantly supervised models, we argue that distant supervision inevitably introduces incomplete and noisy annotations, which may mislead the model training process. To address this issue, we propose a robust NER model named BOND-MoE based on Mixture of Experts (MoE). Instead of relying on a single model for NER prediction, multiple models are trained and ensembled under the Expectation-Maximization (EM) framework, so that noisy supervision can be dramatically alleviated. In addition, we introduce a fair assignment module to balance the document-model assignment process. Extensive experiments on real-world datasets show that the proposed method achieves state-of-the-art performance compared with other distantly supervised NER.