Figure 7From: Semi-supervised learning of causal relations in biomedical scientific discourse Argument location results. Self-training results for the argument location when varying τ and the seed size.Back to article page