Detection of fake news in the Syrian war.

Abstract

After almost eight years of conflict, the humanitarian situation in Syria continues to deteriorate year after year. With multiple opposing parties involved in the armed conflict, much of the news reported about the Syrian war seems to be biased or inclined to support a certain party over the others. With serious human rights violations taking place in the Syrian war, and news sources blaming different sides of the conflict for these violations, interest in the detection of fake news surrounds the Syrian war. In this work, we built a streaming and scraping model to extract news articles of interest from news sources' websites. We built a labeled dataset of news articles about the Syrian conflict. Finally, we built a feature extraction model along with a machine learning model that is able to detect fake news in the Syrian conflict and generalize to other types of fake news.