| 520 |
|
|
|a Radicalisation is a process that historically used to be triggered mainly through social interactions in places of worship, religious schools, prisons, meeting venues, etc. Today, this process is often initiated on the Internet, where radicalisation content is easily shared, and potential candidates are reached more easily, rapidly, and at an unprecedented scale (Edwards and Gribbon, 2013; Von Behr et al., 2013). In recent years, some terrorist organisations succeeded in leveraging the power of social media to recruit individuals to their cause and ideology (Farwell, 2014). It is often the case that such recruitment attempts are initiated on open social media platforms (e.g., Twitter, Facebook, Tumblr, YouTube) but then move onto private messages and/or encrypted platforms (e.g., WhatsApp, Telegram). Such encrypted communication channels have also been used by terrorist cells and networks to plan their operations (Gartenstein-Ross and Barr). To counteract the activities of such organisations, and to halt the spread of radicalisation content, some governments, social media platforms, and counter-extremism agencies are investing in the creation of advanced information technologies to identify and counter extremism through the development of Artificial Intelligent (AI) solutions (Correa and Sureka, 2013; Agarwal and Sureka 2015a; Scrivens and Davies, 2018). These solutions have three main objectives: (i) understanding the phenomena behind online extremism (the communication flow, the use of propaganda, the different stages of the radicalisation process, the variety of radicalisation channels, etc.), (ii) automatically detecting radical users and content, and (iii) predicting the adoption and spreading of extremist ideas. Despite current advancements in the area, multiple challenges still exist, including: (i) the lack of a common definition of prohibited radical and extremist internet activity, (ii) the lack of solid verification of the datasets collected to develop detection and prediction models, (iii) the lack of cooperation across research fields, since most of the developed technological solutions are neither based on, nor do they take advantage of, existing social theories and studies of radicalisation, (iv) the constant evolution of behaviours associated with online extremism in order to avoid being detected by the developed algorithms (changes in terminology, creation of new accounts, etc.) and, (v) the development of ethical guidelines and legislation to regulate the design and development of AI technology to counter radicalisation. In this book chapter we provide an overview of the current technological advancements towards addressing the problem of online extremism (with a particular focus on Jihadism). We identify some of the limitations of current technologies, and highlight some of the potential opportunities. Our aim is to reflect on the current state of the art and to stimulate discussions on the future design and development of AI technology to target the problem of online extremism
|
