General and Domain Adaptive Chinese Spelling Check with Error Consistent Pretraining
This work solves the lack of labeled data and domain adaptation issues in Chinese Spelling Check, which is crucial for improving text accuracy in applications like input methods, but it is incremental as it builds on existing token classification and dictionary methods.
The paper tackles the problem of Chinese Spelling Check (CSC) by developing ECSpell, which uses an Error Consistent masking strategy for pretraining to bridge the gap between automatically generated and real-world data, outperforming previous state-of-the-art models on general benchmarks. It also addresses domain adaptation by adding a user dictionary module (ECSpell$^{UD}$), which significantly improves performance on domain-specific datasets, approaching general benchmark levels.
The lack of label data is one of the significant bottlenecks for Chinese Spelling Check (CSC). Existing researches use the method of automatic generation by exploiting unlabeled data to expand the supervised corpus. However, there is a big gap between the real input scenario and automatic generated corpus. Thus, we develop a competitive general speller ECSpell which adopts the Error Consistent masking strategy to create data for pretraining. This error consistency masking strategy is used to specify the error types of automatically generated sentences which is consistent with real scene. The experimental result indicates our model outperforms previous state-of-the-art models on the general benchmark. Moreover, spellers often work within a particular domain in real life. Due to lots of uncommon domain terms, experiments on our built domain specific datasets show that general models perform terribly. Inspired by the common practice of input methods, we propose to add an alterable user dictionary to handle the zero-shot domain adaption problem. Specifically, we attach a User Dictionary guided inference module (UD) to a general token classification based speller. Our experiments demonstrate that ECSpell$^{UD}$, namely ECSpell combined with UD, surpasses all the other baselines largely, even approaching the performance on the general benchmark.