This is what a graph of 8,000 fake Twitter accounts looks like

Abstraction: Network graph analysis revealing 8,000 interconnected Twitter bot accounts

Key points:

• Bots use randomly generated usernames (e.g., "ekip_uhokoqeq"), hijacked bios and profile photos from real users, and repost random tweets
• Crawl used a Python/NetworkX script to 3–5 levels of follower graph depth; Twitter API rate limits make deeper crawls slow
• Each bot follows ~8 others; graph shows clustering and "loop" structures (closed networks of bots following only each other)
• Visualized with Gephi, Cytoscape (Allegro and Organic layouts), and Python Graph-Tool
• Loop centers appeared to be real accounts; one central node had 650,000 followers, likely mostly fake
• Author speculates bots evade Twitter filters by forming internal-only networks before pivoting to spam or being sold as fake followers

Connections: Twitter · Gephi · Social Network Analysis · Graph Visualization · Bot Detection

Source: https://shkspr.mobi/blog/2015/03/this-is-what-a-graph-of-8000-fake-twitter-accounts-looks-like/