CSMR researchers describe a method for inferring network structure using different similarity metrics for social media data. Rather than focusing on agreement between different scores, the researchers use differences and disagreements to capture unique structures and behaviors. Their findings help platforms and researchers understand previously-hidden structures in the development of online communities.
Srayan Datta, University of Michigan Chandra Phelan, University of Michigan Eytan Adar, University of Michigan Many social media systems explicitly connect individuals (e.g., Facebook or Twitter); as a result, they are the targets of most research on social networks. However, many systems do not emphasize or support explicit linking between people (e.g., Wikipedia or Reddit), and even fewer explicitly link communities. Instead, network analysis is performed through inference on implicit connections, such as co-authorship or text similarity. Depending on how inference is done and what data drove it, different networks may emerge. While correlated structures often indicate stability, in this work we demonstrate that differences, or misalignment, between inferred networks also capture interesting behavioral patterns. For example, high-text but low-author similarity often reveals communities “at war” with each other over an issue or high-author but low-text similarity can suggest community fragmentation. Because we are able to model edge direction, we also find that asymmetry in degree (in-versus-out) co-occurs with marginalized identities (subreddits related to women, people of color, LGBTQ, etc.). In this work, we provide algorithms that can identify misaligned links, network structures and communities. We then apply these techniques to Reddit to demonstrate how these algorithms can be used to decipher inter-group dynamics in social media. Additional Key Words and Phrases: Inter-group similarity; Reddit; social network analysis; misaligned linksPACM on Human-Computer Interaction, Vol. 1, No. CSCW, Article 37. Publication date: November 2017. https://doi.org/10.1145/3134672 .