"The highlighted tokens are predominantly Turkish morphemes, roots, and suffixes that form the core of words, especially those indicating tense, plurality, possession, or case. Many activations occur at the boundaries of word stems and suffixes, or at the start of compound or derived words, reflecting the agglutinative structure of Turkish. There is also a focus on function words and key content words that carry the main semantic load in a sentence."
Score Type | Accuracy | Precision | Recall | F1 score | TPR | TNR | FPR | FNR |
---|---|---|---|---|---|---|---|---|
detection | 0.94 | 1.0 | 0.88 | 0.936 | 0.88 | 1.0 | 0.0 | 0.12 |
fuzz | 0.93 | 0.939 | 0.92 | 0.929 | 0.92 | 0.94 | 0.06 | 0.08 |