Skip to main content

Table 5 Clustering performance obtained by merging controlled vocabularies, concept mapping and no vocabulary indexing

From: Gene prioritization and clustering by multi-view text mining

Merging vocabulary

RI

NMI

merge-9

0.6321 ± 0.0038

0.2830 ± 0.0079

merge-4

0.6333 ± 0.0053

0.2867 ± 0.0085

concept-4

0.6241 ± 0.0056

0.2644 ± 0.0111

novoc

0.5630

0.0892

  1. The merge-9, merge-4 and concept-4 profiles were clustered by K-means in 20 random repetitions and the mean values and deviations of evaluations are shown in the table. The novoc profile was only evaluated once by K-means because of the extremely high dimension and the computational burden.
Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us