Religious hate speech in the Arabic Twittersphere is a notable problem that requires developing automated tools to detect messages that use inflammatory sectarian language to promote hatred and violence against people on the basis of religious affiliation. Distinguishing hate speech from other profane and vulgar language is quite a challenging task that requires deep linguistic analysis. The richness of the Arabic morphology and the limited available resources for the Arabic language make this task even more challenging. To the best of our knowledge, this paper is the first to address the problem of identifying speech promoting religious hatred in the Arabic Twitter. In this work, we describe how we created the first publicly available Arabic dataset annotated for the task of religious hate speech detection and the first Arabic lexicon consisting of terms commonly found in religious discussions along with scores representing their polarity and strength. We then developed various classification models using lexicon-based, n-gram-based, and deep-learning-based approaches. A detailed comparison of the performance of different models on a completely new unseen dataset is then presented. We find that a simple Recurrent Neural Network (RNN) architecture with Gated Recurrent Units (GRU) and pre-trained word embeddings can adequately detect religious hate speech with 0.84 Area Under the Receiver Operating Characteristic curve (AUROC).

Growing evidence suggests that YouTube's recommendation algorithm plays a role in online radicalization via surfacing extreme content. Radical Islamist groups, in particular, have been profiting from the global appeal of YouTube to disseminate hate and jihadist propaganda. In this quantitative, data-driven study, we investigate the prevalence of religiously intolerant Arabic YouTube videos, the tendency of the platform to recommend such videos, and how these recommendations are affected by demographics and watch history. Based on our deep learning classifier developed to detect hateful videos and a large-scale dataset of over 350K videos, we find that Arabic videos targeting religious minorities are particularly prevalent in search results (30%) and first-level recommendations (21%), and that 15% of overall captured recommendations point to hateful videos. Our personalized audit experiments suggest that gender and religious identity can substantially affect the extent of exposure to hateful content. Our results contribute vital insights into the phenomenon of online radicalization and facilitate curbing online harmful content.

Arabic Twitter space is crawling with bots that fuel political feuds, spread misinformation, and proliferate sectarian rhetoric. While efforts have long existed to analyze and detect English bots, Arabic bot detection and characterization remains largely understudied. In this work, we contribute new insights into the role of bots in spreading religious hatred on Arabic Twitter and introduce a novel regression model that can accurately identify Arabic language bots. Our assessment shows that existing tools that are highly accurate in detecting English bots don't perform as well on Arabic bots. We identify the possible reasons for this poor performance, perform a thorough analysis of linguistic, content, behavioral and network features, and report on the most informative features that distinguish Arabic bots from humans as well as the differences between Arabic and English bots. Our results mark an important step toward understanding the behavior of malicious bots on Arabic Twitter and pave the way for a more effective Arabic bot detection tools.

YouTube strives to moderate its content by censoring, demonetizing or removing videos that allegedly violate their community guidelines. Such strategies, especially if seen as unjust by the affected users, could be met with resentment, anger, and in some cases, violence. In addition to YouTube removing videos, uploaders sometimes delete their videos for a variety of reasons such as paraphrasing or preserving online self-image. In this paper, we provide a detailed analysis of deleted/removed videos on YouTube. To do this, we tracked over 73,000 recent YouTube videos for one week and identified those that got deleted or removed. We have then conducted a large-scale analysis of this data and reported on the most informative features that distinguish deleted/removed videos from the ones that remain available. Based on our analysis, we have developed machine learning prediction models that predict videos that will get deleted/removed at different stages of a video's lifetime, viz., at the time of posting, and after up to seven days have elapsed. Our findings indicate that we can predict video deletion/removal with high accuracy even at the time of posting—a strategy that could help users perceive the removal of their videos as fair as well as reduce public and moderators exposure to problematic videos.