. Author manuscript; available in PMC: 2014 Jun 1.

Published in final edited form as: J Biomed Inform. 2013 Feb 27;46(3):436–443. doi: 10.1016/j.jbi.2013.02.001

Table 4.

Results of the different clustering methods for Radiology reports data.

Method descriptor	Clustering algorithm	Data modality utilized	Micro-Averaged Precision^* (95% confidence interval)	NMI ^* (95% confidence interval)
KmeansWords	Kmeans	Words	0.510 (0.506, 0.514)	0.236 (0.234, 0.238)
KmeansConcepts	Kmeans	Concepts	0.560 (0.555, 0.565)	0.468 (0.447, 0.490)
EnsembleWords	Ensemble	Words	0.551 (0.534, 0.567)	0.211 (0.204, 0.218)
EnsembleConcepts	Ensemble	Concepts	0.591 (0.588, 0.594)	0.565 (0.551, 0.579)
EnsembleMerged	Ensemble of merged modalities	Words and concepts merged	0.589 (0.584, 0.593)	0.487 (0.485, 0.489)
Complementary Ensemble Clustering	Linear combination of coassociation matrices from ensemble clustering of each modality separately	Words and concepts separately	0.609 (0.599, 0.620)	0.550 (0.540, 0.560)

NMI: Normalized Mutual Information.

Bold font indicates highest performing method(s).

Larger is better.