Research on Knowledge Graph Completion Model Combining Temporal Convolutional Network and Monte Carlo Tree Search

Figure 6

Positive rate of rewards of Policy-UCT search on each dataset (C = 1.96).