Contribution Classification Methods for Fake News Using Machine Learning

Based on the review of articles, it can be inferred that online social networks have evolved into a significant and influential platform for communication, exchange of ideas, and sharing information. However, they also facilitate the rapid and widespread dissemination of misinformation, which may lead to adverse effects on individuals or society. The World Wide Web contains data in various formats such as documents, videos, and audio files. Detecting and classifying online news in unstructured formats (such as news articles, videos, and audio files) is relatively challenging as it requires human expertise. In this regard, the aim of the present research was to review the classification methods for fake news using machine learning. Through this, reputable domestic and foreign articles were reviewed, and it was ultimately determined that methods such as geometric deep learning, maximum entropy, deep neural network algorithms, semi-supervised learning, and others are effective for detecting and classifying fake news.