Fake News Detection Using Ensemble Approach

This paper introduces an ensemble-based technique that uses Natural Language Processing method to detect fake news. It uses Support Vector Machine, Decision Tree, Passive Aggressive Classifier, and Artificial Neural Network, four different types of classifiers. The research's dataset for the study consisted of news pieces with labels indicating their truthfulness. Pre-processing of the data included text cleaning, tokenization, and stemming. Using the initially processed data, the classifiers were trained and evaluated, and their performance were measured using metrics such as F1 score, accuracy, recall, as well as precision. The findings revealed that ensemble model with a voting classifier outperformed individual classifiers, achieving maximum accuracy. The findings suggest that the proposed method for identifying false news is effective and has the potential to benefit various fields such as journalism, media, politics, and public policy. The usage of ensemble techniques for detecting fake news has gained popularity in recent years, and this paper highlights a novel method that employs NLP and machine learning techniques.