Skip to main content

Understanding how emotions can spread articles throughout a social network

Imagine opening up your Facebook/Twitter/Instagram feed and checking out what’s going on. A given front page might consist of status updates from friends, groups you belong to sharing invite links to events, or even articles and news being taken in through either sponsored content/advertising or (more interestingly) other people you are connected to sharing these articles and content to friends in their own networks.

However, what compels someone to, when seeing an article, press the “share” button?

This article analyzes how content spreads throughout a network, more specifically how emotions can encourage people to share individual articles. The paper above, more specifically, uses a dataset of all New York Times articles published over a three-month period to examine exactly how emotion, and the emotional content of an article, ties into exactly how viral an article would become. More specifically, this paper has found that articles that elicit strong emotions from viewers (such as those which elicit feelings of awe, anger, or happiness) will often be more likely to become viral: this explains many of the trends involved in viral marketing, such as titles like “7 things about (subject) that will BLOW your MIND,” or articles which, as they spread, make people more and more angry about individual issues and thus more likely to share these articles. In fact, articles that elicit anger are 34% more likely to stay on the front page of the New York Times: they’re more popular because they create more discussion about the content that they report on. The article also analyzes the reverse, hypothesizing that deactivating emotions (such as sadness) in articles would make people less likely to share those articles: based on their findings from the NYT dataset, they have proven this as well. Understanding this underlying trend is necessary to understand how articles that evoke “active emotion” (anxiety, anger, awe) will inevitably be more viral than articles that don’t, and this ties greatly into what opinions are read about, talked about, and ultimately acted on by a larger majority of the population.

This greatly ties into the information we have learned in class as well, more specifically the likelihood of information cascades occurring based on a given article. While we have focused mostly on what would happen when, given a certain set of parameters and initial “converts” of an article, a given network would eventually adopt the article in full, we’ve not necessarily considered where these parameters have come from. This article aims to understand the sociological and psychological trends which cause articles to, for instance, only require nodes to have 1/2 of their neighbors be converted for them to be converted or convinced to share the article as well, or to have perhaps two or three specific nodes in specific clusters of a network want to spread it. Consider, for instance, a political article. In a given network, it might be the case that if a few nodes in separate networks (potentially of different political alignments) share a given article, the article will eventually spread to be shared by the entire network based on the given parameters we’ve seen in class. However, what decides those parameters is the article’s contents: if the article is a particularly energizing one, the bar for actually spreading this article throughout a network would be significantly lower. The above paper essentially extends our understanding of what goes on in class primarily by giving a potential cause to why social media posts and shared and reshared, spreading throughout a network.

Considering that someone tweets a link to a New York Times story once every four seconds, and as 59% of people report that they frequently share online content with others, it’s pretty easy to see that sharing online content is an integral part of modern life: in today’s day and age, one’s opinions and beliefs are shaped intensively by what one sees online, especially if content online is shared by friends or close connections. Understanding how opinions are shaped, and the implications of emotions in what articles become viral is important to understand how any individual article or series of events will ultimately impact us as a society.



Leave a Reply

Blogging Calendar

November 2018