Fake news detection for Arabic headlines-articles news data using deep learning

Fake news has become increasingly prevalent in recent years. The evolution of social websites spurred the expansion fake causing it to a mixture with truthful information. English detection had largest share studies, unlike Arabic detection, which is still very limited. phenomenon changed people and perspectives through revolts several Arab countries. False results distortion reality ignite chaos stir public judgments. This paper provides an approach using different deep learning models including long short-term memory convolutional neural network based on article-headline pairs differentiate if headline fact related or unrelated parallel article. In this paper, dataset created about war Syria Middle East political issues utilized. whole data comprises 422 claims 3,042 articles. yield promising results.