Forged News Recognition Using Ensemble Machine Learning Classification Techniques

Forged news broadcast by media outlets poses a real threat to the credibility of information, catching forged news that has become increasingly popular in recent years. Most forged news is deliberately written to mislead readers, making it very difficult to detect forged news based solely on news content. At the same time, forged news will have real evidence to criticize real news, revealing different levels of deception and increasing the difficulty of investigation. On the other hand, forged news creates different types of information from different perspectives. This comprehensive database provides a rich overview of forged news and provides unprecedented access to forged news. In this article, forged detection of forged news with different levels of forged news by combining multiple sources has been developed. Specifically, this model provides a method for combining data from different sources and distinguishing the forged news level, and we offer a multisite, multiclass forged news detection framework, which combines automatic extraction, multisite fusion, and automation level. Experimental results on real world data demonstrate the effectiveness of the proposed method, and further experiments were undertaken to gain a deeper understanding.