Text Mining of Trade War in Indonesia News (Tempo.co): A Wordcloud, Sentiment Analysis, and Cluster

This study focuses on analysing news headlines from Indonesia's leading English daily newspaper, Tempo.co. with total of 4,184 news headlines. We were manually collected it from April to June 2025. The data was processed and analysed using the R Studio package for text mining and sentiment analysis. Various methods such as tokenisation, standardisation, data cleaning, stopword removal, stemming, and lemmatisation were used in the pre-processing stage to extract information. The research methodology included techniques such as wordcloud, sentiment analysis, and clustering to identify the most frequently occurring words, emotional tones, and groups of words that are interrelated in news headlines. Based on the results of the text mining analysis, it was found that the majority of news headlines focused on President Donald Trump's speech on Liberation Day and its impact on trade policy, particularly regarding import tariffs on trading partner countries. This coverage influenced public perception and business decision-making in political and economic aspects.