Comparing the Effectiveness of Different Machine Learning and Deep Learning Models for Fake News Detection

Abstract

Today's information ecosystem, with its abundance of fake news, presents serious obstacles to accurate reporting. Several machine learning and deep learning models have been implemented to help in the detection of fake news. In this research, we evaluate the merits and weaknesses of various algorithms for spotting false news. Deep learning models LSTM and CNN are examined alongside popular machine learning models such as Gaussian Naive Bayes, Decision Tree Classifier, Random Forest Classifier, XGBoost, and LightGBM. Accuracy is the metric against which the models are measured. In terms of accuracy in classifying false news stories, the results demonstrate that the deep learning models, and in particular LSTM and CNN, perform better than the machine learning models. While CNN is good at capturing structural information and local relationships, LSTM excels at capturing long-term dependencies and language patterns. The research emphasises the superiority of deep learning models in detecting false news, providing useful insights for the creation of trustworthy detection systems. The results pave the way for more studies into false news detection methods and contribute to their development.