Skip to main content

Table 2 Performance of systems on the clinical STS corpus of the BioCreative/OHNLP shared task in 2018

From: Distributed representation and one-hot representation fusion with gated network for clinical semantic textual similarity

MethodScore Interval
[0,1][1, 2][2, 3][3, 4][4, 5]Overall
Baseline
 One-hot0.55670.23110.09980.24090.11670.7939
 CNN0.3960−0.0850−0.00900.0370−0.06540.4444
 LSTM0.3920−0.29450.2088−0.0538−0.03030.4275
 BERT0.76130.12060.26350.25300.12100.8461
Concatenation
 CNN + one-hot0.54060.69170.13520.25390.07440.8083
 LSTM+one-hot0.58500.34150.22690.21730.21550.8030
 BERT+one-hot0.66840.30380.23090.24250.22030.8525
Fusion (gated network)
 CNN + one-hot0.69730.23240.16750.23360.08640.8442
 LSTM+one-hot0.62530.35830.18690.25500.10180.8379
 BERT+one-hot0.68720.16050.32380.28220.16660.8541