Table 2 Results for the BT, GETM, I2B2 and OSIRIS data sets using the Int.NP.Sim. model with a confidence threshold of 95% for mean ambiguity reduction ( ÎĽAmb.Red. ), mean recall ( ÎĽ Recall ), and the harmonic mean of mean ambiguity reduction and recall ( H ( ÎĽAmb.Red., ÎĽRecall ))
From: Generalising semantic category disambiguation with large lexical resources for fun and profit
Data set | ÎĽ A m b . R e d . | ÎĽ R e c a l l | H( ÎĽ A m b . R e d ., ÎĽ R e c a l l) |
|---|---|---|---|
BT | 78.00/+34.00 | 99.54/-00.31 | 87.46/+26.38 |
GETM | 88.50/+32.50 | 99.99/-00.01 | 93.89/+22.10 |
I2B2 | 77.60/+42.60 | 98.14/-01.50 | 86.67/+34.87 |
OSIRIS | 78.00/+42.00 | 99.79/-00.21 | 87.56/+34.62 |