CharBERT: Character-aware Pre-trained Language Model
This work addresses the issue of out-of-vocabulary words and misspellings in NLP models, offering enhanced robustness for applications like text classification and sequence labeling, though it is incremental as it builds on existing models like BERT.
The paper tackles the problem of incomplete and fragile word representations in pre-trained language models by proposing CharBERT, a character-aware model that fuses character and subword representations, resulting in significant improvements in performance and robustness on tasks like question answering and text classification.
Most pre-trained language models (PLMs) construct word representations at subword level with Byte-Pair Encoding (BPE) or its variations, by which OOV (out-of-vocab) words are almost avoidable. However, those methods split a word into subword units and make the representation incomplete and fragile. In this paper, we propose a character-aware pre-trained language model named CharBERT improving on the previous methods (such as BERT, RoBERTa) to tackle these problems. We first construct the contextual word embedding for each token from the sequential character representations, then fuse the representations of characters and the subword representations by a novel heterogeneous interaction module. We also propose a new pre-training task named NLM (Noisy LM) for unsupervised character representation learning. We evaluate our method on question answering, sequence labeling, and text classification tasks, both on the original datasets and adversarial misspelling test sets. The experimental results show that our method can significantly improve the performance and robustness of PLMs simultaneously. Pretrained models, evaluation sets, and code are available at https://github.com/wtma/CharBERT