Implementasi Metode Term Frequency-Inverse Document FrequencyClass Frequency untuk Peringkasan Berita Online

Marwa Sulehu<br> Juhar<br> Watty Rimalia<br> Akbar iskandar (STMIK AKBA)

Celebes Engineering Journal
Vol 1 No 1 (2019): Celebes Engineering Journal

This study aims to improve and implement the Term Frequency Inverse Document Frequency Method with the Class Frequency method in summarizing online news to save time for news readers in understanding news through news summaries. Data (text documents) used in this study amounted to 20 Indonesian language news documents obtained from the site http://www.kompas.com. The trial document is a collection of news from the economic sports and technology categories. Data were analyzed using the Term Frequency Inverse Document Frequency Class Frequency method. The results of this study, show that Class Frequency implementation can affect the accuracy of word weighting in the Term Frequency and Inverse Document Frequency methods where the test results obtain an average accuracy of up to 75% of the 20 documents tested by comparing system testing with manual testing.