An Extended Sequence Tagging Vocabulary for Grammatical Error Correction
This is an incremental improvement for natural language processing applications like writing assistance.
The paper tackled the problem of grammatical error correction by extending a sequence-tagging approach with specialized tags for spelling and inflection, resulting in improved performance on the BEA benchmark.
We extend a current sequence-tagging approach to Grammatical Error Correction (GEC) by introducing specialised tags for spelling correction and morphological inflection using the SymSpell and LemmInflect algorithms. Our approach improves generalisation: the proposed new tagset allows a smaller number of tags to correct a larger range of errors. Our results show a performance improvement both overall and in the targeted error categories. We further show that ensembles trained with our new tagset outperform those trained with the baseline tagset on the public BEA benchmark.