Research Article

Biomedical Text Classification Using Augmented Word Representation Based on Distributional and Relational Contexts

Table 4

Word similarity performance with and .

Word embeddingsBioSimLexBioSimVerbMENMTurkRG65RWSCWSSimLex999TR9856UMNSRS-relUMNSRS-simVERB143WS353WS353RWS353S

GloVe_W0.414470.035000.112180.25121−0.084050.193610.344900.109360.146810.082760.160100.257750.239490.158270.27814
GloVe_C0.407110.0341040.113190.25104−0.071610.194330.349300.106570.148610.080630.157380.249710.244060.16950.27649
GloVe_Merged0.422320.030790.110850.24424−0.070910.195340.345760.103820.147250.082240.169110.238340.242840.163550.29231
WE0.431590.130040.198730.287950.136090.215290.376970.152180.156060.258850.338190.282550.342540.274830.40136
CE0.452620.133120.200810.285090.189090.178720.372750.154490.151260.229890.324020.284310.348230.312350.40330
Merged0.432540.174970.218340.276770.221490.134590.376110.152990.158170.282890.353500.186350.288670.259550.32509

Bold means the best performance in the case of each dataset.