TextClassification

model	MRR	MAP	Precision1
BertCAFE	0.56	0.52	0.37
BertDPCNN(Conv2D)	0.56	0.530	0.37
BertDPCNN(conv1D)	0.57	0.537	0.38
BertLSTM	0.58	0.54	0.40
BertHAN	0.581	0.545	0.384
BertRCNN	0.59	0.55	0.41
3输入BERT+CNN+cat	0.63	0.58	0.46
3输入BERT+两种交互+CNN+cat	0.627	0.578	0.457
3输入BERT+TripleAtt+CNN+cat	0.6326	0.5837	0.4639
3输入BERT+TripleAtt+RNN+CNN+cat	0.6357	0.5862	0.4684
3输入BERT+TripleAtt+联合Att+meanmaxPooling	0.6348	0.5854	0.4639
3输入BERT+TripleAtt+Transformer+meanmaxPooling	0.6410	0.5941	0.4714
3输入BERT+TripleAtt+CNN+meanmaxPooling	0.6448	-	-