Journal of Big Data

Table 9 K-means clustering AC (WebKB, tf-normal)

From: Pairwise document similarity measure based on present term set

	K = 5	K = 10	K = 15	K = 20	K = 25	K = 30
PDSM	0.5941	0.5989	0.5552	0.6060	0.5745	0.5670
Manhattan	0.4172	0.4296	0.3933	0.4025	0.3947	0.3942
Euclidean	0.5704	0.5893	0.6053	0.6090	0.6251	0.6263
Cosine	0.5667	0.5937	0.6047	0.6013	0.6086	0.6103
EJ	0.5463	0.5754	0.5909	0.5915	0.6160	0.6162

Maximum accuracy is italicized

Back to article page