PageRank and Facebook
https://www.searchenginejournal.com/facebook-spam/273823/
This article describes how networks of accounts and pages affect advertisement and fake news on Facebook. There is a lot of research on Search Engine Optimization and how Google reconfigures its algorithm to weed out people who manipulate SEO. However, there is a new term that has come about in recent years – Social Media Optimization. Essentially, social media optimization increases traffic to websites based on engagement metrics and meaningful content. Content that has many shares, likes, comments, etc (engagement) is rewarded and bumped higher on peoples’ feeds.
We talked about PageRank in class and the effect of Search Engine Optimization on our search results. Essentially, a website’s importance is determined by the number and quality of links to it. Theoretically, the more and better quality the links, the more important or relevant the website because it is referenced often in a useful way. This same mentality applies to Social Media Optimization. Just as there are articles on people “hacking” Google SEO by finding innovative ways to interlink their content even though they may not be naturally useful, Facebook has also run into this issue with spammers.
Spammers used Facebook’s algorithm to drive up their engagement metrics. They would post political content that would be sure to generate discussion on their Facebook Pages and in hundreds of groups, and as this would surely result in a lot of likes, comments, shares, etc, they now had high engagement metrics. With these high metrics, they get more and more traffic to their sites (something Facebook calls Ad farms), and earn more money based on this high traffic.
This loop shows one of the dangers of using systems similar to PageRank – deceptive information can easily become widespread as spammers find ways to take advantage of the algorithms. Creating hundreds of fake pages and accounts is relatively easy for bots, and this creates a network of deception which negatively influences how people act as well as cheats Facebook’s advertisement model. As I was reading this, I wondered how heavily this related to what occurred in the Presidential Election. Were these spammers a large part of the skewed content on Facebook and is this how they manipulated the algorithm?
Facebook has recently began looking at behavior in addition to meaningful content to mitigate this problem – by adding a metric of looking at whether fake accounts are being used or if there is repeated posting of spam, Facebook can better detect which pages and groups to remove. Nonetheless, while they are making progress on finding and removing spammers, we should all be cautious and aware of signs that can signify if we are on an Ad farm page or a regular page because our thoughts and actions can be manipulated otherwise based on heavily slanted and potentially manipulative content.