Using Information Cascades as an Approach to Fake News
Information Cascade Mechanism and Measurement of Indonesian Fake News by Andry Alamsyah et al: https://ieeexplore-ieee-org.proxy.library.cornell.edu/stamp/stamp.jsp?tp=&arnumber=9527415
This article aims to take a look at the role social media actors play in fake news, particularly in Indonesia, and develop an information cascade model to identify instances of it. In this particular article, Alamsyah et al found that an information cascade taking place online begins most often with a tweet about a topic and that tweet gets circulated via the “retweet mechanism.” Alamsyah et al took retweet data and were able to categorize the fake news topics into three categories: government, politics, and health. They used a Social Networks Analysis Model (SNA) to “model the relationship between node and edge, which represent actor and their relationship” (Alamsyah et al, 2021) and a Susceptible-Infected Model (SI), which is a model used to classify the “stage of disease affecting each individual”, with distinguished individuals into one of two groups: “Susceptible (S(t)) is for individuals who have not received and retweeted tweets that have fake news elements at a given time (t), while infectious (I(t)) individuals are those who have [29]” (Alamsyah et al, 2021). They concluded that the spread of fake news reaches more people and occurs more quickly than true news.
This article takes what we’ve covered in class about information cascades and applies it to reality. We’ve seen a rise of fake news in our society over the past few years and it’s something to be modeled and measured so that we can better know how to recognize and combat it. The beginning of an information cascade in this context was a tweet, which was spread essentially by retweets, resulting in a wide circulation of the piece of “news.” In a lot of ways, this is similar to the herding experiment with the red and blue urns. Both begin with a piece of information: the tweet or the ball you’ve drawn. Then, the information is announced, or circulated: the tweet gets published or the guess is made public. Following this, other people begin spreading that information and making decisions based on it: the tweets get retweeted and people begin to believe them, or the sequential decisions are influenced by what the people ahead of them have to say. To me, the most astounding part of it all is the power that the individual holds to break the fragile information cascade with the urns, which begs the question: is it harder to break a fake news information cascade as opposed to the one described in the herding experiment? In my opinion, it can be. The stakes can feel higher when dealing with fake news, and what I’ve seen happen is entire belief systems shaped around fake news circulations. Once these beliefs are solidified, it seems no amount of science or retraction can break them. This makes it all the more important to continue to develop models and mechanisms to combat fake news every chance we get, because as they say: knowledge is power—which makes corrupted information all the more dangerous.