Automatic Text Summarization of Madura Tourism Articles Using TF-IDF and K-Medoid Clustering

Abstract

Tourism is one of the most popular article topics. Reading articles takes time, especially if you read many article documents. An automatic text summary system offers a solution to shorten the time reading articles. Text summaries are composed of essential sentences from documents that are less or more than half in length. The length of the summary result amount is determined by the summarizer by specifying a threshold manually. This study proposes that TF-IDF and K-Medoid perform Extractive automatic text summarization of Indonesian language documents. This method is expected to automatically determine summary result limits and get good accuracy in automatic text summarization. The TF-IDF algorithm is known to be able to generate text summaries based on the weights obtained for each sentence in the text. The K-Medoid method is a clustering method that functions to group sentences considered by the system to have the same meaning. This study uses data totaling 33 document texts taken from different web sources on the internet. To test the study’s summary results, expert respondents will manually correct the summary results from the system. From the results of the tests that have been carried out, the accuracy value is 65.40%.