Table 2 Translation evaluation based on final coding results
From: Automatic RadLex coding of Chinese structured radiology reports based on text similarity ensemble
Method | Precision (%) | Recall (%) | F1-score (%) |
|---|---|---|---|
Google translation + MLP weighting text similarity ensemble algorithm | 59.33 | 81.61 | 68.71 |
Dictionary translation + MLP weighting text similarity ensemble algorithm | 87.42 | 77.27 | 82.03 |
Hybrid translation + MLP weighting text similarity ensemble algorithm | 91.78 | 88.59 | 90.15 |
Human translation + MLP weighting text similarity ensemble algorithm | 97.26 | 90.37 | 93.69 |