Klasterisasi Minat Global Terhadap Konflik Iran–Israel Menggunakan K-Means Pada Data Google Trends (13–24 Juni 2025)

Abstract

The conflict between Iran and Israel is one of the geopolitical issues that has attracted the most global attention in the past decade. Entering an open escalation phase since June 13, 2025, the conflict has triggered a surge in online information searches from people around the world. In this context, public search behavior can reflect collective concerns and be an important indicator in reading global perceptions of an international crisis. This study aims to group countries in the world based on their search patterns for the Iran-Israel conflict using a Google Trends-based data mining approach. The study was conducted by collecting Google Trends data for the first 12 days of the conflict escalation, using three main keywords: “Iran Israel conflict”, “Iran Israel war”, and “Iran Israel News”. The method used is K-Means Clustering, an unsupervised learning algorithm that is effective in forming clusters based on the similarity of data patterns. Data from more than 100 countries were processed through preprocessing, normalization, clustering, and visualization stages using Principal Component Analysis (PCA). The analysis results showed that there were three main classes that represented differences in global search characteristics. The first cluster was dominated by countries with high interest in war aspects, the second class reflected equal interest in all keywords, and the third class showed low search intensity overall. Countries with the highest number of search interests included Afghanistan, Albania, and Algeria. It was concluded that bold search data can be used as an indicator of global attention to conflict issues. This approach is relevant to quickly understand public perception patterns, and can be applied to other international issues in digital geopolitical studies.