STUDI KOMPARATIF ALGORITMA K-MEANS CLUSTERING DAN FUZZY C-MEANS CLUSTERING UNTUK PENGELOMPOKAN DATA ARTIKEL ILMIAH

Maryama Kurnia Amri, NIM.: 22206051001 (2025) STUDI KOMPARATIF ALGORITMA K-MEANS CLUSTERING DAN FUZZY C-MEANS CLUSTERING UNTUK PENGELOMPOKAN DATA ARTIKEL ILMIAH. Masters thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (STUDI KOMPARATIF ALGORITMA K-MEANS CLUSTERING DAN FUZZY C-MEANS CLUSTERING UNTUK PENGELOMPOKAN DATA ARTIKEL ILMIAH) 22206051001_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version Download (6MB) | Preview
	Text (STUDI KOMPARATIF ALGORITMA K-MEANS CLUSTERING DAN FUZZY C-MEANS CLUSTERING UNTUK PENGELOMPOKAN DATA ARTIKEL ILMIAH) 22206051001_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf - Published Version Restricted to Registered users only Download (12MB) | Request a copy

Abstract

In recent times, the annual number of published scientific articles has risen to millions. This significant increase poses challenges for researchers, particularly when attempting to understand clustering trends in the development of studies within specific fields. Therefore, tools capable of automating the clustering process for scientific articles are essential. This study aims to evaluate algorithms that are potentially suitable for clustering scientific article data. Specifically, this study aims to compare K-Means and Fuzzy C-Means (FCM) algorithm’s performance for clustering scientific article data. An experimental method was employed, using a dataset of 7,136 articles comprising titles and abstracts obtained from various digital libraries, including IEEE, ACM, and Springer Link. Based on the Silhouette Coefficient scores, for title attributes across four tested keywords, K-Means outperformed FCM on two keywords, while the remaining two keywords showed equal scores for both algorithms. For abstract attributes, FCM performed better on three out of four keywords. Furthermore, the Wilcoxon test on survey responses revealed no significant differences in the clustering results between the two algorithms. The findings indicate that K-Means excels in clustering title attributes, while FCM is more effective for abstract attributes based on Silhouette scores. However, according to respondent opinions, the clustering results between K-Means and Fuzzy C-Means (FCM) exhibit no significant differences

Share this knowledge with your friends :

Actions (login required)

View Item