TiSpell: A Semi-Masked Methodology for Tibetan Spelling Correction covering Multi-Level Error with Data Augmentation

TL;DR

TiSpell is a novel semi-masked Tibetan spelling correction model that effectively handles multi-level errors using data augmentation, outperforming existing methods on both simulated and real-world datasets.

Contribution

We introduce a semi-masked model and a data augmentation strategy for Tibetan spelling correction addressing multi-level errors, filling gaps in open datasets and improving correction accuracy.

Findings

TiSpell outperforms baseline models in experiments.

Our data augmentation improves model robustness.

TiSpell matches state-of-the-art performance.

Abstract

Multi-level Tibetan spelling correction addresses errors at both the character and syllable levels within a unified model. Existing methods focus mainly on single-level correction and lack effective integration of both levels. Moreover, there are no open-source datasets or augmentation methods tailored for this task in Tibetan. To tackle this, we propose a data augmentation approach using unlabeled text to generate multi-level corruptions, and introduce TiSpell, a semi-masked model capable of correcting both character- and syllable-level errors. Although syllable-level correction is more challenging due to its reliance on global context, our semi-masked strategy simplifies this process. We synthesize nine types of corruptions on clean sentences to create a robust training set. Experiments on both simulated and real-world data demonstrate that TiSpell, trained on our dataset, outperforms baseline models and matches the performance of state-of-the-art approaches, confirming its effectiveness.