Network Analysis to Discover Emerging Influencers & Superspreaders

To overcome the limited availability of social network platform data, we can use the NetworkX Python package to uncover opinion leaders by tracking and analyzing real-time sharing data as a temporal network in an iterative process. At each time interval, the network graph grows as new nodes are identified and added based on how often their posts are being shared; additionally, we calculate social media influencers’ centrality score to rank their influence and reveal emerging threats to the community.

This workshop will train trust and safety analysts to apply social network analysis (SNA) to identify producers and superspreaders of messages. No previous knowledge of Python or SNA required. We provide an overview of network modeling, data structures, and analysis concepts. Then, we learn NetworkX to construct, quantify, visualize relationships and interactions. We conclude by discussing how to build the technical infrastructure within your organization to capture signals and surface alerts.

Stanford University, Trust & Safety Conference

September 2024