"The highlighted tokens are primarily stems, roots, or affixes of words in Portuguese (and some in Spanish, French, German, and Hindi), often marking verb conjugations, noun/adjective forms, or place names. These tokens frequently appear at morpheme boundaries, within inflected or derived forms, and in multiword proper nouns or locations, reflecting the morphological structure and multilingual context of the text."
Score Type | Accuracy | Precision | Recall | F1 score | TPR | TNR | FPR | FNR |
---|---|---|---|---|---|---|---|---|
detection | 0.7 | 0.647 | 0.88 | 0.746 | 0.88 | 0.52 | 0.48 | 0.12 |
fuzz | 0.67 | 0.608 | 0.96 | 0.744 | 0.96 | 0.38 | 0.62 | 0.04 |