Article Per Year (5 Year)
p-Index From 2020 - 2025
0
P-Index
This Author published in this journals
All Journal Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Novia Agusvina
Fakultas Ilmu Komputer, Universitas Brawijaya
Computer Science & IT Control & Systems Engineering Education Electrical & Electronics Engineering Engineering

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Pengelompokan Artikel Berbahasa Indonesia Dengan Menggunakan Reduksi Fitur Information Gain Thresholding Dan K-Means Novia Agusvina; Indriati Indriati; Nurudin Santoso
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 10 (2018): Oktober 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (568 KB)

Abstract

The increasing number of articles spread on the internet site, making it difficult for users to find the desired article. One of the online article service providers is Kompas.com. To face the competition among mass media industry, Kompas.com step is to provide features that facilitate the user, such as features related article recommendations. However, in its application Kompas.com is still less than the maximum so it remains inferior to other online mass media. In this study, researchers implemented a method of reducing the features of Information Gain Thresholding and K-Means to create a group of related articles. The purpose of this study is to improve the system related articles from Kompas.com. In implementing the use of java language. In the early stages of preprocessing to reduce the disturbance in the data, then the feature reduction is done to reduce the features used for faster process, then weighted as the basis for calculating the distance between documents, after finding the distance of the initial distance or centroid, grouping can be done. The results show that the clustering of articles using Information Gain Threshold and K-Means is good enough, has criteria of silhouette coefficient of 0.9595 and a purity measure of 0.75 with 3 clusters and 0.04 threshold limit, this conclude that it gives better purity compared to without feature reduction.
Co-Authors Indriati Indriati Nurudin Santoso