Table 1.
Details comparison of the related literature of news categorization.
Fig 1.
Methodology of news classification and sentiment analysis.
Fig 2.
Category of news with news frequency.
Fig 3.
Flowchart of sentiment analysis.
Fig 4.
Top-down approach of news categorization using ML.
Both supervised and unsupervised algorithms are used, and individual performance is analyzed.
Fig 5.
An illustration of NMF.
Fig 6.
Block diagram of proposed blending SGDR algorithm.
Fig 7.
Polarity and subjectivity of BBC news dataset.
Table 2.
Sentiment analysis.
Table 3.
Hyperparameters of the classifiers and the suitable values obtain by GridSearchCV.
Fig 8.
Range of the classification metrics for individual news category using ML classifiers.
Table 4.
Performance of distance-based algorithms to news categorization.
Table 5.
Performance of probability-based algorithms to news categorization.
Table 6.
Performance of tree-based algorithms to news categorization.
Table 7.
Performance of optimization and regularization method to news categorization.
Fig 9.
Range of the classification metrics for individual news category using unsupervised ML classifiers.
Table 8.
Performance of unsupervised algorithms to news categorization.
Fig 10.
Comparison of the accuracy of the algorithms before and after string processing.
Table 9.
Performance of the classifiers after applying the string preprocessed techniques.
Fig 11.
Tops words for each news category.
Table 10.
McNemar test result of proposed SGDR compared to other algorithms.
Table 11.
10-Fold cross-validation accuracy of the algorithms.
Table 12.
Ablation study.