Research Article
Deep Learning Approaches for Cyberbullying Detection and Classification on Social Media
Table 4
Results of existing with the proposed SSA-DNN on predicting cyberbullying with 90% of training data.
| Measures | NB | LR | RF | SVM | ANN | ANN-DRL | SSA-DBN |
| Accuracy | 99.503 | 99.633 | 99.623 | 99.673 | 99.563 | 99.703 | 99.983 | F-measure | 88.071 | 88.191 | 90.213 | 90.063 | 91.454 | 91.455 | 94.378 | G-mean | 96.299 | 96.269 | 96.530 | 96.640 | 97.030 | 97.000 | 98.361 | Sensitivity | 92.796 | 92.766 | 93.537 | 93.697 | 94.387 | 94.418 | 96.037 | Specificity | 99.673 | 99.583 | 99.773 | 99.803 | 99.733 | 99.873 | 99.997 |
|
|