Özet:

Abstract Nowadays, Due to the growth of online communication, many people use social media platforms and produce a lot of content. Fake news creates negative perceptions of society. The rise of social media networks has given fake news a platform to rapidly gain popularity among users. Identifying and labelling Arabic fake news represents a big challenge because of the large amount of heterogeneous content in addition to the limited related Arabic datasets available. machine learning (ML), Natural language processing (NLP) and deep learning (DL) are commonly used to increase the speed and automate the analytical process of this huge amount of content and to transform unstructured text into structured form. In this study, a corpus of news websites is developed to determine fake news using some machine learning techniques; this includes a dataset 3185 fake news articles and a new dataset consists of 1453 real news articles. This paper shows that, using an aggregation of machine learning and ensemble methods, we can make a prediction model for fake news that has an accuracy of up to 100%. with low complexity, which can save power and energy, and we can use it as a reference for the detection of fake news in Egypt and Arab countries.

Anahtar Kelimeler: