List your company
ResearchFormerly WizWand

Joint Word Segmentation and POS Tagging on Japanese (test)

90.8F1 ScoreViterbi
51.38461.61771.8582.083May 31, 2025
Updated 8mo ago

Evaluation results

MethodLinks
ViterbiEncoder layers=2 layers, Training approach=CRF training2025.0590.8
ViterbiEncoder layers=4 layers, Training approach=CRF training2025.0590.7
ViterbiEncoder layers=4 layers, Training approach=BCRF training2025.0590.7
ViterbiEncoder layers=No self-attentive encoder, Training approach=CRF training2025.0590.6
BCRFEncoder layers=4 layers, Training approach=BCRF training, Iterations=102025.0590.6
ViterbiEncoder layers=2 layers, Training approach=BCRF training2025.0590.5
BCRFEncoder layers=2 layers, Training approach=CRF training, Iterations=102025.0590.4
BCRFEncoder layers=2 layers, Training approach=BCRF training, Iterations=102025.0590.4
BCRFEncoder layers=4 layers, Training approach=CRF training, Iterations=102025.0590.2
BCRFEncoder layers=No self-attentive encoder, Training approach=CRF training, Iterations=102025.0590.1
BCRFEncoder layers=4 layers, Training approach=BCRF training, Iterations=52025.0590.1
ViterbiEncoder layers=No self-attentive encoder, Training approach=BCRF training2025.0589.8
BCRFEncoder layers=2 layers, Training approach=BCRF training, Iterations=52025.0589.8
BCRFEncoder layers=No self-attentive encoder, Training approach=BCRF training, Iterations=102025.0589.6
MFEncoder layers=4 layers, Training approach=MF training, Iterations=102025.0589.3
MFEncoder layers=No self-attentive encoder, Training approach=MF training, Iterations=102025.0589.1
MFEncoder layers=2 layers, Training approach=MF training, Iterations=102025.0589.1
UnstructuredEncoder layers=4 layers2025.0589
MFEncoder layers=4 layers, Training approach=MF training, Iterations=52025.0589
MFEncoder layers=2 layers, Training approach=MF training, Iterations=52025.0588.8
BCRFEncoder layers=No self-attentive encoder, Training approach=BCRF training, Iterations=52025.0588.8
BCRFEncoder layers=No self-attentive encoder, Training approach=CRF training, Iterations=52025.0588.7
BCRFEncoder layers=2 layers, Training approach=CRF training, Iterations=52025.0588.7
MFEncoder layers=No self-attentive encoder, Training approach=MF training, Iterations=52025.0588.7
BCRFEncoder layers=4 layers, Training approach=CRF training, Iterations=52025.0588.5
UnstructuredEncoder layers=2 layers2025.0586.3
ViterbiEncoder layers=4 layers, Training approach=MF training2025.0583.5
ViterbiEncoder layers=2 layers, Training approach=MF training2025.0582.3
ViterbiEncoder layers=No self-attentive encoder, Training approach=MF training2025.0579.6
MFEncoder layers=No self-attentive encoder, Training approach=CRF training, Iterations=102025.0577.6
MFEncoder layers=No self-attentive encoder, Training approach=CRF training, Iterations=52025.0576.5
MFEncoder layers=2 layers, Training approach=CRF training, Iterations=102025.0576.1
MFEncoder layers=4 layers, Training approach=CRF training, Iterations=102025.0576
MFEncoder layers=2 layers, Training approach=CRF training, Iterations=52025.0575.4
MFEncoder layers=4 layers, Training approach=CRF training, Iterations=52025.0575.1
UnstructuredEncoder layers=No self-attentive encoder2025.0552.9