High Efficiency Spam Filtering: A Manifold Learning-Based Approach

Table 2

Different algorithms’ classification accuracy.

Datasets	SVM [19]	SVM [18]	SVM + LEP	SVM + LEP + Struc
EnronSpam	92.1%	93.5%	93.9%	94.7%
PU1	95.8%	96.1%	95.6%	96.9%
GenSpam	93.7%	94.6%	93.9%	95.1%