%A NIM.: 18106050051 Najmunda Zia Akmal
%O Pembimbing: Maria Ulfah Siregar, S.Kom. MIT., Ph.D.
%T KOMPARASI ALGORITMA NAIVE BAYES DAN K-NEAREST
NEIGHBORS UNTUK KLASIFIKASI JUDUL BERITA
%X The existence of news that has existed for a long time makes the amount of
produced news very large and will continue to grow. In addition, news themes are
also very diverse, where almost everything can be discussed in the news. The
variety of themes makes news-making companies have to group the news, which
is still done manually. The application of machine learning algorithms can be a
solution to group news automatically.
This study aims to compare the value of accuracy, precision, recall, f1,
AUC, and Train-Prediction Time of the Naive Bayes and KNN algorithms in the
classification of news titles that consist of 4 news categories.
` The results showed that the Naive Bayes algorithm worked better than KNN
with accuracy value, f1, AUC, and Train-Prediction Time are 0.840, 0.840, 0.964,
0.090 second versus 0.796, 0.797, 0.944, 0.944 second for data with one word as
token; 0.851, 0.851, 0.968, 0.107 second versus 0.801, 0.802, 0.945, 0.927 second
for data with 1 and 2 words as tokens; and 0.853, 0.853, 0.968, 0.115 second
versus 0.795, 0.796, 0.944, 1.095 second for data with 1, 2, and 3 words as
tokens.
%K classification; naive bayes; KNN; News
%D 2022
%I UIN SUNAN KALIJAGA YOGYAKARTA
%L digilib53631