Table 7 Result of argument analysis using hierarchical attention network (Word Embedding from Scratch)
From: Argument annotation and analysis using deep learning with attention mechanism in Bahasa Indonesia
No | Batch size | Accuracy (%) | Recall (%) | Precision (%) | F1 macro (%) | ROC-AUC |
|---|---|---|---|---|---|---|
1 | 16 | 72.69 ± 3.44 | 61.36 ± 6.10 | 67.51 ± 17.20 | 59.69 ± 10.29 | 81.23 ± 4.13 |
2 | 32 | 74.84 ± 3.01 | 64.98 ± 5.13 | 77.19 ± 3.47 | 65.17 ± 6.36 | 84.66 ± 1.69 |
3 | 64 | 77.46 ± 3.54 | 69.28 ± 5.76 | 78.65 ± 3.74 | 70.29 ± 6.76 | 86.16 ± 2.90 |
4 | 100 | 69.79 ± 5.19 | 58.98 ± 11.14 | 49.04 ± 19.74 | 52.87 ± 16.15 | 73.73 ± 7.86 |
5 | 128 | 69.76 ± 4.77 | 56.05 ± 8.10 | 50.91 ± 21.83 | 50.06 ± 13.19 | 75.68 ± 7.33 |