Forecasting Political Security Threats: A Fusion of Lexicon-Based and ML Approaches

This paper focuses on monitoring online sentiments and opinions to enhance national security. Excessive emotions expressed online can potentially lead to threats like riots and civil unrest, which jeopardize social and political stability. Researchers highlight the connection between emotions, sentiments, and political security risks. To address this, this paper introduces a novel framework that predicts political security threats using a hybrid approach combining lexicon-based analysis and machine learning in cyberspace. The decision tree, Naive Bayes, and support vector machine classifiers are also employed. Experimental analysis demonstrates that the hybrid lexicon-based approach with decision tree achieves the highest performance in predicting political security threats, emphasizing the framework’s effectiveness.