Skip to main content

Table 9 Comparison results (%accuracy) on progress notes

From: A multitask bi-directional RNN model for named entity recognition on Chinese electronic medical records

Model Entity type
  Disease Symptom Disease group Treatment Test Overall accuracy
Naive Bayes (NB) 69.50 70.09 N/A 41.59 71.85 67.49
Maximum Entropy (ME) 71.49 72.37 41.15 52.93 77.58 72.44
Support Vector Machine (SVM) 77.77 76.92 21.12 56.36 81.49 76.45
Conditional Random Field (CRF) [7] 87.42 87.09 36.06 75.60 90.31 87.22
Convolutional Neural Network(CNN) [7] 76.19 76.65 12.50 51.83 76.65 73.40
Bi-RNN model 87.48 87.01 25.00 63.99 83.75 82.72
Transfer learning Bi-RNN model [24] 88.70 88.49 31.25 72.93 86.12 85.43
Our proposed model 92.24 94.19 75.00 86.46 92.61 92.13