Lena Clever, Tim Schatz-Eckrodt, Lena Frischlich and colleagues conducted a case study on the German group Generation Islam to examine how they use Instagram to spread Islamic extremist content.
The researchers examined 1187 posts that were collected over a period of 2 years, from January 2016 to December 2018.
They examined affect in hashtag networks in which users may come across propagandistic content, used deep learning to examine the emotional valence of the visuals and employed automated linguistic analysis to describe the collective action cues contained within the texts.
In general, it has to be noted that Islamic propaganda spread via social media is more common than one might think. The team extracted number of public posts using the specific hashtag from Instagram, and, using inductive coding, the hashtags were sorted into eight different categories. Next, an image analysis was conducted with Imavis. SentiBank was deployed to detect distinct emotions. For the linguistic analysis, the LIWC dictionary was used.
The team found that the most frequent 15 hashtags could be sorted into the category of religion, but a combination of religious and non-religious hashtags was most commonly used. The hashtags relied on positive affect.
Imavis classified most pictures as negative, and the analysis with SentiBanks confirmed these findings (anger, fear, defense-related images dominated). The textual elements anchored the social identities of users who were attracted by the religious hashtags and mostly spread a positive view of the future if users followed “the right path”.
Find the full article here: https://journals.sagepub.com/doi/epdf/10.1177/20563051221150404