Intrusion Detection Techniques in Social Media Cloud: Review and Future Directions

Table 4

AI and ML method review table.

Reference	Methodology	Dataset	Result
[33]	Naïve Bayesian classifier and a decision tree	KDD CUP 99	99% detection rate
[34]	Naïve Bayesian classifier	KDD CUP 99	The accuracy for detecting normal traffic was 100% with a false positive of 0.03%, the accuracy of detecting probing activities was 99.95% with a false positive of 0.36%, the accuracy of detecting DoS was 100% with a false positive of 0.03%, the accuracy of detecting U2R was 99.67% with a false positive of 0.10%, and the accuracy of detecting R2L was 99.58% with a false positive of 6.71%
[35]	SVM	NSL-KDD	The accuracy for 23 features of NSL-KDD was 99.32, whereas the accuracy for 30 features was 99.37
[36]	Hybrid model that incorporates data mining approaches such as the -means clustering algorithm and the RBF kernel function of the support vector machine	KDD CUP 99	The proposed model KMSVM has an accuracy of 92.86 percent
[37]	Decision tree (J48) algorithm	KDD CUP 99	The study result shows that there are 97.23% correctly classified instances, whereas there is 2.67% of incorrectly classified instances with a high true-positive rate of 99% for normal and attack packets
[38]	RF	KDD CUP 99	Accuracy of the proposed model was 96.42% with a false-positive rate of 0.98%
[39]	C4.5 decision tree	NSL-KDD	The study result shows that the decision tree-based intrusion detection system has 75.7 accuracy, 0.6 false-positive rate, and 0.95 true-positive rates, and the time for building the model was 4.42 seconds
[40]	-means and -nearest neighbors	KDD CUP 99	Proposed model has good accuracy, and the training time of the proposed model was very suitable
[41]	GA and BFSS	NSL-KDD	The study result shows that by conducting the cross-validation comparison between the GA-BFSS of all features and the three GAs of the feature set, the GA-BFSS has 93.26 whereas the GA1, GA2, and GA3 have 92.12, 92.64, and 93.00, respectively.
[42]	RF and SVM	NSL-KDD	The study result shows that, after selecting the important features, the accuracy of the SVM classifier using all features was 90% for DoS, 89% for Probe, 79% for R2L, and 100% for U2R. In contrast, the accuracy of the RF classifier using all features was 85% for DoS, 88% for Probe, 78% for R2L, and 100% for U2R