KLASTERISASI KEMIRIPAN JUDUL BERITA PALSU PADA BERITA DARING MENGGUNAKAN METODE FUZZY C-MEANS
Total View This Week0
Total View This Week0
Institusion
Institut Teknologi Perusahaan Listrik Negara
Author
HASSAN, MUHAMAD ALWI
Arianto, Rakhmat
Putra, Rakhmadi Irfansyah
Arianto, Rakhmat
Putra, Rakhmadi Irfansyah
Subject
Teknik Informatika
Datestamp
2022-09-21 07:49:06
Abstract :
This study discusses the making of the application of the clustering of Similar News Title Clusters based on the News Title dataset sourced from turnbackhoax.id. to identify online news titles that are false or valid using the Fuzzy C-Means (FCM) method. Fuzzy C-Means has a high degree of accuracy and fast computing time. Fuzzy C-Means can be used to classify data based on certain attributes. Before in to the FCM, the pre-processing stage with the textmining method is to convert text data into numerics to make it easier when used as input parameters into the FCM method which includes tokenizing, filtering, and stemming. Then do the calculation of the number of occurrences of words / term queries in the document being tested and used as input parameters for the calculation of the FCM method. In this study, the Fuzzy C-Means algorithm will be used to cluster the similarity of fake news titles on scattered online news. From the results of the cluster that has the highest value becomes an additional parameter to identify hoaks or valid news. The testing method uses accuracy by calculating what percentage of documents are similar to the news headline input and have an accuracy rate of 64%.
This study discusses the making of the application of the clustering of Similar News Title Clusters based on the News Title dataset sourced from turnbackhoax.id. to identify online news titles that are false or valid using the Fuzzy C-Means (FCM) method. Fuzzy C-Means has a high degree of accuracy and fast computing time. Fuzzy C-Means can be used to classify data based on certain attributes. Before in to the FCM, the pre-processing stage with the textmining method is to convert text data into numerics to make it easier when used as input parameters into the FCM method which includes tokenizing, filtering, and stemming. Then do the calculation of the number of occurrences of words / term queries in the document being tested and used as input parameters for the calculation of the FCM method. In this study, the Fuzzy C-Means algorithm will be used to cluster the similarity of fake news titles on scattered online news. From the results of the cluster that has the highest value becomes an additional parameter to identify hoaks or valid news. The testing method uses accuracy by calculating what percentage of documents are similar to the news headline input and have an accuracy rate of 64%.
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
