Jurnal IEEE : Indonesian news classification based on NaBaNA

Abstract:

This paper focused on the classification of Indonesian news categories. News articles have the format of text, so it will be more complex and needs to be a process to prepare the data. Also, the article is accepted Indonesia language articles should be simplified into a basic word on every word, this can be done by the method of stemmer Nazief and Andriani. For the classification method used is Naïve Bayes method is commonly used for text mining. Both of these methods Naïve Bayes and Nazief-Adriani stemming (NaBaNA) will collaborate to get results with high accuracy. The results showed by Naïve Bayes classification with the support of Nazief and Andriani get higher accuracy.

I. Introduction

Today the development of science and information technology is very rapid; many innovations created information technology to help alleviate human work. Not a few companies or agencies that get big profits after optimizing the utilization of information technology owned. The Internet is one of the media information today that always experience a very rapid development. With the internet one can easily do activities such as reading news, watching movies, finding information, even doing transactions though. Therefore, the internet is considered become source information as background knowledge that required to be classified [1].

Download paper PDF