Research Article
A Case-Based Reasoning Approach for Automatic Adaptation of Classifiers in Mobile Phishing Detection
Table 4
Accuracy comparison of classifiers on 10 features.
| ā | Feature sets | J48 (%) | DT (%) | IBK (%) | LR (%) | NB (%) | SVM (%) |
| 1 | Android components | 93.23 | 89.02 | 93.40 | 90.16 | 84.67 | 87.95 | 2 | API count | 95.85 | 93.02 | 95.66 | 91.90 | 89.20 | 85.25 | 3 | APIusage_actions | 95.20 | 91.86 | 95.32 | 91.97 | 89.02 | 91.24 | 4 | Flow | 93.05 | 91.03 | 93.32 | 87.18 | 87.45 | 83.17 | 5 | Hardware components | 89.00 | 89.06 | 89.12 | 89.06 | 89.02 | 89.06 | 6 | Intent_action | 86.89 | 85.73 | 87.13 | 84.64 | 83.75 | 85.53 | 7 | Permission | 94.30 | 91.92 | 94.65 | 93.95 | 88.54 | 94.14 | 8 | Shell_command_strings | 75.40 | 71.18 | 74.08 | 70.28 | 68.74 | 70.22 | 9 | Content_visual | 97.20 | 95.79 | 95.53 | 94.49 | 95.77 | 93.87 | 10 | URLs | 96.03 | 93.24 | 97.18 | 93.99 | 92.98 | 93.80 |
|
|