Identifying Daesh-Related Propaganda Using OSINT and Clustering Analysis

The development of the digital society has substantially altered the conditions under which conflicts occur. Emerging threats are characterized by their asymmetry, diversity, and constant change; rapid transmission over the network; near-immediate nature; possibility for unrestricted access; and swift ability to alter the behaviour of individuals. This paradox is an example of cognitive warfare, which employs both traditional and novel information, cyber, and psychological warfare techniques. The self-proclaimed Islamic State engages in a unique type of disruptive cyber cognitive-intelligence activity utilizing cyberspace. We now refer to the Weaponization of Media Narratives: the struggle of narratives has overtaken the relevance of traditional military and physical Jihad. Jihadist activities consist of sending threatening messages to Western nations and promoting online propaganda in order to recruit new members and instil terror in individuals. Daesh’s propaganda output is so extensive that it is practically impossible for humans to analyse it. Thus, it is crucial to establish and implement cyber defence strategies to prevent, identify, and deter jihadist Internet activity. Law Enforcement, Intelligence, and other organizations are constantly devising new tools to prevent, identify, and restrict terrorist operations over the Internet. The collection and analysis of information from a vast array of sources can give intelligence analysts with useful insights by revealing previously concealed but logically sound patterns and connections. Beginning with a review of Al-Naba’s propaganda materials, this study seeks to construct an automated model that would aid in detecting and identifying the online locations of Daesh. We looked at Al-Naba’ magazine instead of another newspaper because it has only been published in Arabic. Other magazines have been published in other languages and have been looked at in a lot of community identification and Social Network propaganda analysis studies in the past. Therefore, the purpose of our study was to discover if it is possible to employ computer assistance to evaluate Jihadist tales in order to identify any (thematic) similarities across various propaganda sources. One of the specific goals was to evaluate whether or not there are tweets with a direct connection to Al-Naba’ magazine. We wanted to make sure that the tweets were coded in a way that was consistent with the Twitter data—collected from Kaggle—we used as a training set. This was important because tweets could be put into different groups. This was done to see if the tweets were correctly put into their own groups based on information from Al-Naba’s writings. So, the number of times each group shows up depends on how often it shows up in more than 1% of the texts in each cluster.