This post was written by Carina Albrecht, PhD Candidate and Data Fluencies fellow at the Digital Democracies Institute
On April 12, 2023, we had the pleasure of hosting Dr. Garth Davies for our Spring Speakers Series, who shared a bit of his crucial and timely work with us. Dr. Davies is an Associate Professor in the School of Criminology at SFU, and his research focuses on understanding violent extremism and radicalization, including developing tools for better detecting and predicting violent and extremist activities online.
Dr. Davies is at the forefront of many projects that focus on developing new methods and tools for collecting and analyzing textual data from different places on the web where extremists convene to post and chat, including social media platforms. For example, in collaboration with the RCMP and Computer Science scholars, he helped develop the “Dark-Risk Assessment Framework Tool using Machine Learning” (DRAFT-ML) which uses machine learning not only to detect online violent speech and toxicity but also tries to predict when such instances will result on violent attacks offline. Another of his projects developed the Terrorism and Extremism Network Extractor (TENE), a web crawler that collects data from extremist forums. This data was later used in sentiment analysis software to identify negative emotions that might influence violent behavior, particularly grievance-based emotions. Such uses of technology for differentiating between people who post violent speech on the internet, and those who will go on to perform a violent act offline, is an extremely challenging task. Dr. Davies recognizes that the state of the technology is still in its infancy, and they are “still not ready to turn everything to the computer”; therefore, there is a need to constantly intervene in the algorithms to verify what they are doing and correct its mistakes.
Recently, the COVID-19 pandemic and its potential to fuel radicalization and extremism was one of Dr. Davies’s sites of investigation. Along with other researchers, Dr. Davies investigated seven different online extremist forums, including right-wing extremists, Incels, left-wing extremists, and jihadists. They collected posts from these forums between January 2020 and April 2020 – a couple of months before and after the WHO declared COVID-19 a pandemic – to compare and understand if the pandemic has fueled violent extremism online. They found a significant increase in posting behavior among violent right-wing extremists and Incel forums after March 2020, while they observed no changes on left-wing and jihadist platforms, revealing that the pandemic affected these groups differently.
Dr. Davies’ slide presentation showing the rise of posts on Incel platforms after COVID-19 was declared a pandemic by the WHO.
This rise of extremist activity only in specific forums raises the question: Do conspiracy theories and online disinformation help fuel narratives of extremism and violence within certain groups? If this is the case, these findings underscore the importance of improving content moderation on social media platforms and developing better tools to identify the entanglements between disinformation, sentiment, and violent extremism. Dr. Davies believes this type of work is already very challenging to be taken by one group of society alone. Future work needs to include the participation of civil society – not only in content moderation regulation but also in creating more digital literacy programs to help educate future generations about mis/disinformation online.
You can read more about Dr. Davies’ critical and fascinating case study on COVID-19-related extremism here.