A multi-label classification of online news using multinomial Naive Bayes

Online news is information about an event or occurrence presented through the internet media. The classification of news labels is usually based on the topic discussed in the news article. However, if a news article has more than one topic, then the news article has multiple labels. The classification of labels on news articles that have more than one label can be done with multi-label classification to provide accurate classification results according to the content of the article. One of the most frequently occurring topics in online news articles is sexual violence. In news articles about sexual violence, there is likely more than one topic about sexual violence, so multi-label classification is needed to facilitate the grouping of online news articles to get the appropriate keywords from the sexual violence case. This research aims to perform multi-label classification using the Multinomial Naive Bayes algorithm with a problem transformation approach, namely Binary Relevance and Label Power-set. The data labels used in this study refer to the types of sexual violence crimes included in Law No. 12 of 2022, which consists of 19 labels. This research uses 2203 data from online news portals detik.com and kompas.com. In this study, multi-label classification using Multinomial Naive Bayes and problem transformation methods, namely Label Power-set on the split data of training and testing of 90:10, can perform multi-label classification with the best accuracy of 60.309%. In future research, recommended to add keywords using query expansion and balancing the amount of data on each label.