"The highlighted tokens are primarily high-activation morphemes, word stems, or grammatical particles in Korean and Turkish, as well as some corrupted or unknown characters, often marking key syntactic or semantic roles within sentences. These tokens frequently appear at word boundaries, as suffixes, or as part of compound words, and are crucial for sentence structure or meaning."
Score Type | Accuracy | Precision | Recall | F1 score | TPR | TNR | FPR | FNR |
---|---|---|---|---|---|---|---|---|
detection | 0.78 | 0.868 | 0.66 | 0.75 | 0.66 | 0.9 | 0.1 | 0.34 |
fuzz | 0.55 | 0.535 | 0.76 | 0.628 | 0.76 | 0.34 | 0.66 | 0.24 |