Skip to main content

Table 1 Sentence-level Macro-\(F_{1}\) score for 2 classes (argumentative, non-argumentative) and for 3 classes (PRO, CON, NON) for the in-topic and cross-topic setups from our previous publication [9]

From: Relational and Fine-Grained Argument Mining

  Method In-Topic Cross-Topic
2‑classes BiLSTM 0.74 0.66
BiCLSTM 0.74 0.70
BiLSTM-KG 0.73 0.68
CAM-Bert 0.80 0.67
TACAM-Bert 0.81 0.80
3‑classes BiLSTM 0.56 0.46
BiCLSTM 0.53 0.47
BiLSTM-KG 0.56 0.47
CAM-Bert 0.73 0.61
TACAM-Bert 0.69 0.64