Table 3

phishGILLNET1--PLSA model performance

Number of topics

Dataset combination1 (8 K public corpus)

Dataset combination2 (400 K public corpus)


Perplexity

Computation time (min)

Perplexity

Computation time (min)


2

523.56

1.65

6742.42

32.31

10

278.81

2.52

4441.96

45.20

25

277.62

3.12

1748.55

52.00

50

274.31

3.63

1593.10

65.00

100

273.33

7.42

1461.78

112.50

200

271.27

15.38

1425.36

185.20


Ramanathan and Wechsler EURASIP Journal on Information Security 2012 2012:1   doi:10.1186/1687-417X-2012-1

Open Data