Social networks have drastically changed how people obtain information. News in social net-works is accompanied by images and videos and thus receives more attention from readers as opposed to traditional ones. Unfortunately, fake news publishers often misuse these advantages to spread false information rapidly. Therefore, the early detection of fake news is crucial. The best way to address this issue is to design an automatic detector based on fake news content. So far, many fake news recognition systems including both traditional machine learning and deep learning models are proposed. Given that manual feature extraction methods are very time-consuming, deep learning methods are the preferred tools. This research aims to enhance the performance of existing approaches by utilizing an ensemble of deep learners based on attention mechanisms. To a great extent, the success of an ensemble model depends on the variety of its learners. To this end, we propose a novel loss function that enforces each learner to attend to different parts of news content on the one hand and obtain a good classification accuracy on the other hand. Also, the learners are built on a common deep feature extractor and only differ in their attention modules. As a result, the number of parameters is reduced efficiently, and the overfitting problem is addressed. Additionally, most research in automatic fake news detection is devoted to fully supervised setting. Given that the generation rate of news in social media is drastic and the labeling of a huge amount of data required by fully supervised models is expensive and time consuming, these models are not beneficial in real applications. To address this limitation, we extend our method for semi-supervised setting using effective augmentations, and a novel distribution-aware pseudo-labeling technique. The proposed augmentations enhance the robustness of learners and prevent overfitting effectively. Diverse learners are utilized to annotate the unlab
