"The highlighted tokens are morphemes, suffixes, or name fragments that are significant in identifying proper nouns, place names, and grammatical forms across multiple languages, especially in Slavic, Romance, and Turkic contexts. These tokens often mark inflections, diminutives, or are parts of multi-token named entities, and are important for language identification, morphological analysis, and named entity recognition."
Score Type | Accuracy | Precision | Recall | F1 score | TPR | TNR | FPR | FNR |
---|---|---|---|---|---|---|---|---|
detection | 0.64 | 0.621 | 0.72 | 0.667 | 0.72 | 0.56 | 0.44 | 0.28 |
fuzz | 0.54 | 0.521 | 0.98 | 0.681 | 0.98 | 0.1 | 0.9 | 0.02 |