Klasifikasi Komentar Terhadap Kebijakan Legalisasi Otomatis Konten Dewasa di Youtube Kids Menggunakan Algoritma Naïve Bayes

muhamad padli haikal (Universitas Muhammadiyah Kota Sukabumi)

KOMPUTEK
Vol 7, No 2 (2023): Oktober

Youtube is one of the most popular entertainment media with many facilities offered both paid and free. Youtube has also released several versions, one of which is specifically for children, namely Youtube Kids. Users can provide input on how to use the application on the Google Play Store. Based on this, this study uses negative and positive comments with the research title "Classification of Comments on the Automatic Legalization of Adult Content Policy on Youtube Kids Using the Naïve Bayes Algorithm", and grouping words using the Clustering Cosine Similarity method with the intention of categories related to the word new policy on YouTube Kids to be more specific. Multinomial Naïve Bayes modeling using a confusion matrix evaluation produces an accuracy of 83% with a True Positive value of 129 and False Positive 50, False Negative 15, True Negative 200.