Research Article
Biomedical Text Classification Using Augmented Word Representation Based on Distributional and Relational Contexts
Table 8
Concept categorization performance with
and
.
| Word embeddings | AP | BLESS | Battig | ESSLI_1a | ESSLI_2b | ESSLI_2c | Ohta-10-bio-words |
| GloVe_W | 0.23631 | 0.23 | 0.11489 | 0.43182 | 0.475 | 0.37778 | 0.41379 | GloVe_C | 0.23383 | 0.22 | 0.1158 | 0.43182 | 0.5 | 0.42222 | 0.41379 | GloVe_Merged | 0.25124 | 0.245 | 0.12426 | 0.54545 | 0.55 | 0.46667 | 0.45689 | WE | 0.27612 | 0.295 | 0.12177 | 0.47727 | 0.575 | 0.37778 | 0.44826 | CE | 0.22886 | 0.32 | 0.11872 | 0.47727 | 0.55 | 0.33333 | 0.43104 | Merged | 0.26119 | 0.28 | 0.12732 | 0.5 | 0.575 | 0.4 | 0.48276 |
|
|
Bold means the best performance in the case of each dataset.
|