A Two-Stage Vietnamese Spelling Correction Pipeline Combining Underthesea and BARTpho

Abstract: Vietnamese spelling correction is challenging due to the language’s rich diacritic system, syllable-based tokenization, and the frequent presence of strict entities in administrative andlegal texts. While sequence-to-sequence models achieve strong correction accuracy, they are proneto over-correction and unintended rewriting under domain shift, which limits their reliability in highstakes applications. In this paper, we propose a deployment-oriented two-stage Vietnamese spellingcorrection pipeline. The first stage performs text normalization and conservative error detection usingUnderthesea, combined with entity masking to preserve rigid identifiers and formatting. The secondstage applies context-aware correction with a BARTpho sequence-to-sequence model, followed bydetector-guided post-processing and iterative masked refinement to control unnecessary edits. Tosupport realistic evaluation, we construct a hybrid dataset that mixes synthetic spelling noise withreal-world errors collected from administrative documents. Experiments against strong multilingualand Vietnamese-specific baselines show that the proposed pipeline achieves high correction accuracy while significantly reducing over-correction. Beyond standard end-to-end metrics, we introducedetection-oriented analyses that explicitly quantify correction behavior at flagged positions, providing clearer evidence of practical safety for real-world deployment.Keywords: Vietnamese spelling correction, controlled text editing, sequence-to-sequence models,Underthesea, BARTpho, administrative text.