Skip to main content

Table 1 Percentage of the accuracy (Acc.) when apply naïve Bayes (NB) to detect topics in OMIM with different learning features.

From: Exploring supervised and unsupervised methods to detect topics in biomedical text

NB

S

S+M

S+M+T

S+M+T+A

M

T+A

GM

GM+M

Acc.

54.4

62.2

63.3

66.4

62.6

65.9

47.4

62.0

  1. S = semantic sypes
  2. M = MeSH terms
  3. T = title
  4. A = Abstract
  5. GM = general MeSH terms