Online Library

This thesis examines extremist Muslim narratives, discourse, and ideologies over the internet by using content analysis to thematically delineate and reconstruct them for the purpose of discovering the argumentation mechanisms through which they become persuasive and appealing. The research problem is that dominant theories in social sciences and popular literature create ‘taken for granted’ inferences that relegate extremist ideologies and narratives to the realm of structural contingencies, psychological pathologies, emotive appeal, manipulated religious ideologies, peculiar and unique rationalities or group dynamics. This thesis hypothesizes instead of the existence of a `logical structure` in extremist Muslim narratives. This logical structure is predicated on rationally persuasive arguments (which employ epistemic and instrumental rationality coupled with inductive/deductive reasoning) that appeal to any rational individual but are ultimately leveraged on for morally wrong end-state choices. Unfortunately much of the counter-narratives today seldom address this logical structure and choose to address the more traditional explanations cited above. Themes and argumentation mechanisms stemming from an examination of extremist Muslim narratives in this study demonstrate the presence and workings of this logical structure.

From its start, the so-called Islamic State of Iraq and the Levant (ISIL/ISIS) has been successfully exploiting social media networks, most notoriously Twitter, to promote its propaganda and recruit new members, resulting in thousands of social media users adopting pro-ISIS stance every year. Automatic identification of pro-ISIS users on social media has, thus, become the centre of interest for various governmental and research organisations. In this paper we propose a semantic-based approach for radicalisation detection on Twitter. Unlike most previous works, which mainly rely on the lexical and contextual representation of the content published by Twitter users, our approach extracts and makes use of the underlying semantics of words exhibited by these users to identify their pro/anti-ISIS stances. Our results show that classifiers trained from words’ semantics outperform those trained from lexical and network features by 2% on average F1-measure.

From its start, the so-called Islamic State of Iraq and the Levant (ISIL/ISIS) has been successfully exploiting social media networks, most notoriously Twitter, to promote its propaganda and recruit new members, resulting in thousands of social media users adopting a pro-ISIS stance every year. Automatic identification of pro-ISIS users on social media has, thus, become the centre of interest for various governmental and research organisations. In this paper we propose a semantic graph-based approach for radicalisation detection on Twitter. Unlike previous works, which mainly rely on the lexical representation of the content published by Twitter users, our approach extracts and makes use of the underlying semantics of words exhibited by these users to identify their pro/anti-ISIS stances. Our results show that classifiers trained from semantic features outperform those trained from lexical, sentiment, topic and network features by 7.8% on average F1-measure.

WhatsApp is the most popular messaging app in the world. Due to its popularity, WhatsApp has become a powerful and cheap tool for political campaigning being widely used during the 2019 Indian general election, where it was used to connect to the voters on a large scale. Along with the campaigning, there have been reports that WhatsApp has also become a breeding ground for harmful speech against various protected groups and religious minorities. Many such messages attempt to instil fear among the population about a specific (minority) community. According to research on inter-group conflict, such `fear speech’ messages could have a lasting impact and might lead to real offline violence. In this paper, we perform the first large scale study on fear speech across thousands of public WhatsApp groups discussing politics in India. We curate a new dataset and try to characterize fear speech from this dataset. We observe that users writing fear speech messages use various events and symbols to create the illusion of fear among the reader about a target community. We build models to classify fear speech and observe that current state-of-the-art NLP models do not perform well at this task. Fear speech messages tend to spread faster and could potentially go undetected by classifiers built to detect traditional toxic speech due to their low toxic nature. Finally, using a novel methodology to target users with Facebook ads, we conduct a survey among the users of these WhatsApp groups to understand the types of users who consume and share fear speech. We believe that this work opens up new research questions that are very different from tackling hate speech which the research community has been traditionally involved in.

The world’s most dangerous jihadists no longer answer to al Qaeda. The terrorists we should fear most are self-recruited wannabes who find purpose in terror and comrades on the Web. This new generation is even more frightening and unpredictable than its predecessors, but its evolution just may reveal the key to its demise.

Security informatics and intelligence computation plays a vital role in detecting and classifying terrorism contents in the web. Accurate web content classification using the computational intelligence and security informatics will increase the opportunities of the early detection of the potential terrorist activities. In this paper, we propose a modified frequency-based term weighting scheme for accurate Dark Web content classification. The proposed term weighting scheme is compared to the common techniques used in text classification such as Term Frequency (TF), Term Frequency-Inverse Document Frequency (TF-IFD), and Term Frequency- Relative Frequency (tf.rf), on a dataset selected from Dark Web Portal Forum. The experimental results show that the classification accuracy and other evaluation measures based on the proposed scheme outperforms other term weighting techniques based classification.