<![CDATA[News is one of the important needs for people in various parts of the world. Through the news, the public can know the various information that is happening in the community such as economic, political, health, criminal, and natural disasters. Increasing information needs every day to make some agencies to be able to present the news quickly, accurately, reliably, and accurately through print and electronic media that can be enjoyed by news readers. Increasing the amount of news gained every day resulted in large data accumulation of text documents either online and offline. This makes it difficult in searching and classifying documents as needed. To simplify the classification of text document news in Indonesian, one of them by using the method of NBC and K-Means Clustering. Where the calculation of NBC done not randomly, while the calculation for K-Means Clustering done randomly in this study the calculation is done 5 time on each document so the accuracy result is less accurate when compared with NBC The highest accuracy result obtained from the research of news classification with nbc method is 50% and the average for the whole of the four documents is 45.33%, while the highest accuracy result of the classification using k_means method is 100% and the overall result is 53.26% with details Document_0 obtained an average yield of 54%, document 1 the average yield of 60%, document 2 the overall average yield of 42.56%, and document 3 the overall average yield of 52.48%.]]>
Copyrights © 2018
