Skip to main content

Table 2 Performance comparison on a real-world breast cancer dataset

From: PASCAL: a pseudo cascade learning framework for breast cancer treatment entity normalization in Chinese clinical text

Model Precision Recall F1
Softmax
Bi-LSTM 0.8171±0.0143 0.8796±0.0264 0.8472±0.0221
Bi-OnLSTM 0.8316±0.0205 0.8978±0.0139 0.8635±0.0152
TCN 0.7135±0.0129 0.8218±0.0231 0.7638±0.0245
GCNN 0.8817±0.0117 0.9016±0.0210 0.8921±0.0124
CRF
Bi-LSTM 0.8682±0.0125 0.8905±0.0238 0.8792±0.0201
Bi-OnLSTM 0.8678±0.0187 0.8952±0.0145 0.8813±0.0168
TCN 0.8486±0.0089 0.9076±0.0214 0.8771±0.0179
GCNN 0.9443±0.0126 0.9628±0.0181 0.9535±0.0094
PASCAL (Softmax + CRF)
Bi-LSTM 0.8931±0.0153 0.9121±0.0183 0.9025±0.0168
Bi-OnLSTM 0.9078±0.0149 0.9348±0.0156 0.9211±0.0175
TCN 0.8744±0.0102 0.9342±0.0192 0.9033±0.0149
GCNN 0.9413±0.0156 0.9770±0.0147 0.9589±0.0054