PSL accuracy. p ¡ {0.05⋆, 0.01†, 0.001‡} with paired bootstrap compared to the best baseline.
. | . | Normative Arguments . | Non-normative Arguments . | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | . | ACC . | AUC . | F1 . | F1sup . | F1att . | F1neu . | ACC . | AUC . | F1 . | F1sup . | F1att . | F1neu . |
1 | Random | 33.5 | 50.2 | 32.6 | 27.8 | 30.1 | 39.9 | 33.4 | 49.9 | 32.5 | 28.7 | 28.8 | 40.0 |
2 | Sentiment | 40.8 | 64.1 | 40.7 | 40.6 | 39.1 | 42.4 | 43.7 | 61.1 | 42.2 | 40.0 | 35.2 | 51.5 |
3 | Text Entail | 51.8 | 61.8 | 36.7 | 12.8 | 30.4 | 67.0 | 52.1 | 62.8 | 38.6 | 18.4 | 31.0 | 66.4 |
4 | PSL (R1–R13) | 54.0‡ | 73.8‡ | 52.1‡ | 47.0‡ | 43.6‡ | 65.7‡ | 57.0‡ | 76.0‡ | 54.0‡ | 50.1‡ | 42.6‡ | 69.3‡ |
5 | ∖ Fact | 55.1‡ | 74.3‡ | 52.4‡ | 47.1‡ | 41.6‡ | 68.4‡ | 58.6‡ | 77.1‡ | 55.1‡ | 50.5‡ | 42.2‡ | 72.7‡ |
6 | ∖ Sentiment | 62.1‡ | 77.6‡ | 57.5‡ | 49.1‡ | 45.8‡ | 77.7‡ | 61.3‡ | 77.8‡ | 56.7‡ | 50.3‡ | 44.1‡ | 75.7‡ |
7 | ∖ Causal | 54.4‡ | 73.1‡ | 52.3‡ | 45.4‡ | 45.4‡ | 66.0‡ | 57.6‡ | 76.1‡ | 54.3‡ | 48.7‡ | 43.4‡ | 70.7‡ |
8 | ∖ Normative | 51.8‡ | 68.6‡ | 49.4‡ | 44.3‡ | 40.4† | 63.4‡ | 54.7‡ | 70.3‡ | 51.4‡ | 47.0‡ | 40.3‡ | 66.8‡ |
9 | ∖ Sentiment + Chain | 61.9‡ | 77.7‡ | 57.7‡ | 49.3‡ | 46.2‡ | 77.6‡ | 61.5‡ | 78.0‡ | 57.2‡ | 50.8‡ | 44.7‡ | 76.1‡ |
(a) Kialo |
. | . | Normative Arguments . | Non-normative Arguments . | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | . | ACC . | AUC . | F1 . | F1sup . | F1att . | F1neu . | ACC . | AUC . | F1 . | F1sup . | F1att . | F1neu . |
1 | Random | 33.5 | 50.2 | 32.6 | 27.8 | 30.1 | 39.9 | 33.4 | 49.9 | 32.5 | 28.7 | 28.8 | 40.0 |
2 | Sentiment | 40.8 | 64.1 | 40.7 | 40.6 | 39.1 | 42.4 | 43.7 | 61.1 | 42.2 | 40.0 | 35.2 | 51.5 |
3 | Text Entail | 51.8 | 61.8 | 36.7 | 12.8 | 30.4 | 67.0 | 52.1 | 62.8 | 38.6 | 18.4 | 31.0 | 66.4 |
4 | PSL (R1–R13) | 54.0‡ | 73.8‡ | 52.1‡ | 47.0‡ | 43.6‡ | 65.7‡ | 57.0‡ | 76.0‡ | 54.0‡ | 50.1‡ | 42.6‡ | 69.3‡ |
5 | ∖ Fact | 55.1‡ | 74.3‡ | 52.4‡ | 47.1‡ | 41.6‡ | 68.4‡ | 58.6‡ | 77.1‡ | 55.1‡ | 50.5‡ | 42.2‡ | 72.7‡ |
6 | ∖ Sentiment | 62.1‡ | 77.6‡ | 57.5‡ | 49.1‡ | 45.8‡ | 77.7‡ | 61.3‡ | 77.8‡ | 56.7‡ | 50.3‡ | 44.1‡ | 75.7‡ |
7 | ∖ Causal | 54.4‡ | 73.1‡ | 52.3‡ | 45.4‡ | 45.4‡ | 66.0‡ | 57.6‡ | 76.1‡ | 54.3‡ | 48.7‡ | 43.4‡ | 70.7‡ |
8 | ∖ Normative | 51.8‡ | 68.6‡ | 49.4‡ | 44.3‡ | 40.4† | 63.4‡ | 54.7‡ | 70.3‡ | 51.4‡ | 47.0‡ | 40.3‡ | 66.8‡ |
9 | ∖ Sentiment + Chain | 61.9‡ | 77.7‡ | 57.7‡ | 49.3‡ | 46.2‡ | 77.6‡ | 61.5‡ | 78.0‡ | 57.2‡ | 50.8‡ | 44.7‡ | 76.1‡ |
(a) Kialo |