Fig 1.
Arabic hate speech detection framework.
Fig 2.
Visualization of dataset sizes.
Table 1.
Example of keywords for data extraction.
Table 2.
Dataset description summary.
Fig 3.
Web portal for manual data labeling.
Fig 4.
Arabic hate speech corpus creation.
Table 3.
Confusion matrix.
Table 4.
Performance results comparison obtained on N-gram feature on Arabic and code-mixing.
Fig 5.
N-gram highest accuracy result visualization.
Fig 6.
TF-IDF highest accuracy result visualization.
Table 5.
Performance results comparison obtained on TF-IDF feature for Arabic and code-mixing.
Fig 7.
CBOW highest accuracy result visualization.
Table 6.
Performance results comparison obtained on CBOW feature for Arabic and code-mixing.
Fig 8.
Comparison of the proposed method with baselines.
Table 7.
Comparison of the proposed approach with the baselines.