Papers with SCoNE
SCoNE: a Self-Correcting and Noise-Augmented Method for Complex Biological and Chemical Named Entity Recognition (2026.eacl-long)
Copied to clipboard
| Challenge: | Named Entity Recognition (NER) is a fundamental task aimed at identifying entities such as names, locations, and organizations. |
| Approach: | They propose a self-correcting and noise-augmented method for complex Biological and Chemical Named Entity Recognition that improves learning diversity and confidence. |
| Outcome: | The proposed method outperforms baseline methods on CHEMDNER and microbial ecology datasets by 1.80 and 2.73 F1-scores. |