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

Table 9

Concept categorization performance with and .

Word embeddings	AP	BLESS	Battig	ESSLI_1a	ESSLI_2b	ESSLI_2c	Ohta-10-bio-words
GloVe_W	0.22388	0.235	0.11871	0.43182	0.525	0.42222	0.5
GloVe_C	0.22139	0.235	0.11527	0.43182	0.525	0.37778	0.49137
GloVe_Merged	0.23383	0.275	0.12349	0.52272	0.55	0.44444	0.49137
WE	0.26368	0.285	0.12005	0.52273	0.525	0.37778	0.47414
CE	0.26866	0.31	0.12388	0.5	0.5	0.4	0.5
Merged	0.25871	0.315	0.12330	0.65909	0.525	0.37778	0.42241

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