Scholarlink

This research proposes a novel hybrid framework for predicting political security threats by integrating a dynamic lexicon-based approach with advanced machine learning techniques. Recognizing the challenges posed by static lexicons and opaque ML models, the system dynamically updates its lexicon using unsupervised learning algorithms and leverages models such as support vector machines, decision trees, and deep learning networks to accurately classify and predict threats. The hybrid framework offers significant improvements in adaptability, interpretability, and predictive accuracy. Through comprehensive evaluation using datasets derived from political unrest events, the model demonstrates superior performance, making it a robust tool for government agencies and analysts to proactively address political threats.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Political Security Threat Prediction Framework Using Hybrid Lexicon Based Approach and Machine Learning Technique

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Shiva Teja Ragula,
  • Pattlola Srinivas

摘要

This research proposes a novel hybrid framework for predicting political security threats by integrating a dynamic lexicon-based approach with advanced machine learning techniques. Recognizing the challenges posed by static lexicons and opaque ML models, the system dynamically updates its lexicon using unsupervised learning algorithms and leverages models such as support vector machines, decision trees, and deep learning networks to accurately classify and predict threats. The hybrid framework offers significant improvements in adaptability, interpretability, and predictive accuracy. Through comprehensive evaluation using datasets derived from political unrest events, the model demonstrates superior performance, making it a robust tool for government agencies and analysts to proactively address political threats.