Multipath Cross Graph Convolution for Knowledge Representation Learning

Table 4

Comparison of baseline models of entity type prediction accuracy.

Dataset	YAGO43kET (%)	FB15kET (%)	WN18RR (%)
TransE	85.36	92.15	75.04
TransE-ET	90.76	93.23	77.13
PTransE	88.53	92.79	78.25
ETE	90.82	94.01	79.31
ConnectE-(E2T+0)	91.78	94.45	81.07
ConnectE-(E2T + TRT)	92.33	94.49	81.22
TransC-GCN_ AdaGrad	91.63	94.61♣	83.27
TransC-GCN_ SGD	92.75♣	94.58	84.55♣