Harnessing AI to Identify and Combat Fake News

Abstract

Fake news is a growing issue, particularly on social media platforms, where approximately 71% of fake news circulates globally. This study explores techniques to differentiate between real and fake news using machine learning classifiers like Decision Trees, MLP, Logistic Regression, Random Forest, and Support Vector Machines (SVM). The analysis includes word distribution, sentence structure, and named entity recognition (NER). Experimental results show that the Random Forest classifier achieves accuracy with 93.14%. Effective preprocessing, such as tokenization and stopword removal, plays a crucial role in improving performance. Despite advancements, challenges like data diversity and fairness remain highlights for future work.

Ähnliche Arbeiten