How to Deeply Analyze the Content of Online Newspapers Using Clustering and Correlation

Abstract

The increase in the number of visitors is one of the keys to increasing income for online newspapers, whether to increase the number of ads, Google AdSense, or customer trust. Therefore, finding which news categories increase the number of visitors needs to be known and analyzed more deeply. Because it is very common to add content to online newspaper sites every day, even for hours, this pattern analysis is not the same as analyzing regular website content patterns. This study intends to add methods in the world of research on how to analyze website content, especially online news, by using the clustering method to classify what news categories bring high, medium, or a low number of visitors and then analyzing the correlation to explore the depth of the relationship between the variables, namely which parameters have a large or low effect on the increase in the number of visitors. A local Batam-based online newspaper company is used as a case study for this research. Data is collected, preprocessed first, and analyzed using the clustering and correlation method. This analysis of the news content readership suggests what news categories should be optimized because it provides an increase in the number of visitors. A summary of the analysis steps in this study is presented. We also provided some suggestions if other online newspaper owners or researchers are interested in a similar analysis of online news content.