Time Frame Detection Based on Online News Documents Using Vector Space Model

Abstract

This study aims at finding an approach to identify news case duration or time frame as well as to identify the ways to find a time frame through the news document. News document used in this research refers to the previous research done by Rahim in 2018. The research uses three popular news website of Indonesia in 2018 which are tribunnews.com, detik.com and liputan6.com. However, in this research, the news website also uses as an object of analysis. The result conducted that the process to find a news time frame should define the cosine similarity minimum (threshold) between news and keyword (query). The researcher also finds that the workflow to determine the time frame uses news document based on the vector space model and combined with measurement technique if-idf and cosine similarity.