End-to-end Recurrent Neural Network Models for Vietnamese Named Entity Recognition: Word-level vs. Character-level
This work addresses named entity recognition for Vietnamese, an incremental improvement as it matches top system performance using deep learning.
The paper tackled Vietnamese named entity recognition by developing end-to-end neural network models, achieving an F1 score of 88.59% on a standard test set without using syntactic or hand-crafted features.
This paper demonstrates end-to-end neural network architectures for Vietnamese named entity recognition. Our best model is a combination of bidirectional Long Short-Term Memory (Bi-LSTM), Convolutional Neural Network (CNN), Conditional Random Field (CRF), using pre-trained word embeddings as input, which achieves an F1 score of 88.59% on a standard test set. Our system is able to achieve a comparable performance to the first-rank system of the VLSP campaign without using any syntactic or hand-crafted features. We also give an extensive empirical study on using common deep learning models for Vietnamese NER, at both word and character level.