Collusion Network
Source: https://www.cbsnews.com/news/collusion-network-facebook-flaw-leads-to-millions-of-fake-likes/
The rise of social media allows more opportunities for hackers to manipulate the system for their own gain. Researchers have found that hackers have used a security loophole to allow more than a million Facebook accounts to easily generate over hundred million likes and comments as part of a “a thriving ecosystem of large-scale reputation manipulation.” The main purpose for this hack is to increase posts’ visibility, since posts that quickly gain a lot of likes are higher in people’s feeds. The hackers harness a code called “OAuth,” which allow third-party applications to access people’s Facebook accounts. An application of this hack is to the 2016 presidential election, where researchers tracked hundreds of phony accounts that were allegedly originated in Russia, who used vast networks to propagate false stories and ads. This hack is harmful, not only because it can foster false propaganda and bring high visibility to detrimental articles, but also because these attackers can also steal personal info as well as propagate malware through the collusion networks.
Facebook has said in a statement that it has blocked collusion networks, though other news says that that statement has been unverified. In connection with material we learned in class, we learned a few ways these collusion networks can be detected. These “fake” or spam accounts can be detected by recognizing accounts that communicate with people more randomly. More specifically, these fake accounts would not have satisfy the strong triadic closure property because the accounts they communicated with would not be communicating with each other. In terms of a graph, the fake account would be a single node in the middle, which edges connecting to a million different nodes, but no other edges connection these nodes (i.e. the image below depicting the graph of 8,000 fake twitter accounts). Graphing all of the nodes and links of accounts could likely help researchers to detect patterns in these fake accounts to pinpoint which accounts are spam.
In addition, these collusion networks show a vast and dangerous application of a network, one in which this course hopes to understand. This example of collusion network shows the significance of network theory to better comprehend the world around us.
Image source: https://shkspr.mobi/blog/2015/03/this-is-what-a-graph-of-8000-fake-twitter-accounts-looks-like/