KOMBO represents Korean characters using subcharacter combination rules from Hangeul, outperforming prior state-of-the-art Korean PLMs by an average of 2.11% across five natural language understanding tasks.
InProceedings of the 54th Annual Meeting of the Association for Computational Lin- guistics (Volume 1: Long Papers), pages 1715–1725, Berlin, Germany
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
KOMBO: Korean Character Representations Based on the Combination Rules of Subcharacters
KOMBO represents Korean characters using subcharacter combination rules from Hangeul, outperforming prior state-of-the-art Korean PLMs by an average of 2.11% across five natural language understanding tasks.