Text Restoration on Igbo Text (full dataset)
98.71AccuracyLRCV
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| LRCVModel Type=Logistic Regression Cross-Validation2026.01 | 98.71 | 94.23 | 93.57 | 93.16 | |
| 5-GramModel Type=n-gram, n=52026.01 | 96.27 | 96.61 | 96.84 | 96.5 | |
| IgTrainModel Type=Embedding Model, Training Text=Igbo text2026.01 | 90.44 | 89.13 | 89.43 | 88.95 | |
| MLconfiguration=Best classification model2026.01 | 81.55 | 75.11 | 71.74 | 72.86 | |
| Bestdescription=Combination of the best performing models on the wordkeys2026.01 | 80.05 | 78.19 | 72.21 | 73.31 | |
| N-gramconfiguration=Best model2026.01 | 80.01 | 78.15 | 72.2 | 73.3 | |
| 5-Gramn-gram order=52026.01 | 80.01 | 78.15 | 72.2 | 73.3 | |
| 4-Gramn-gram order=42026.01 | 79.62 | 77.48 | 71.55 | 72.55 | |
| 3-Gramn-gram order=32026.01 | 78.68 | 75.82 | 69.94 | 70.66 | |
| 2-Gramn-gram order=22026.01 | 77.21 | 73.52 | 66.67 | 66.45 | |
| Embconfiguration=Best embedding model2026.01 | 71.24 | 67.85 | 65.51 | 65.19 | |
| BaselineDescription=Comparison of marked text with stripped version2026.01 | 66.99 | 30.79 | 29.53 | 29.72 | |
| Unigramconfiguration=Baseline2026.01 | 66.75 | 32.3 | 48.17 | 38.55 | |
| 1-Gramn-gram order=12026.01 | 66.75 | 32.3 | 48.17 | 38.55 |