Klasifikasi Penyebab Penyalahgunaan Narkoba Dari Berita Online Dengan Menggunakan Naive Bayes

• Main Author: Laili Wahyunita
• Format: Article
• Language: Indonesian
• Published: P3M Politeknik Negeri Banjarmasin 2017-06-01
• Series: Jurnal ELTIKOM: Jurnal Teknik Elektro, Teknologi Informasi dan Komputer
• Subjects: naive bayes; preprocessing data; text document representatio; presicion; recall
• Online Access: http://eltikom.poliban.ac.id/index.php/eltikom/article/view/12

This research conducted the classification process by applying the method of classification of Naive Bayes. News article document is one form of text data that is not structured so that requires the process of cleaning data and pre-processing first. The Naive Bayes approach is an approach that refers to Bayes's Theorem, where it uses the principle of statistical opportunity to combine previous knowledge. The use of this technique is based on the need of the system to know the probability value of the data to be classified. Waterfall method was used for built this classificaiton system. Accuracy rate up to 60 % with the highest precision and recall is 80% and 90%.