Dissecting the Meme Magic: Understanding Indicators of Virality in Image Memes

要旨

Despite the increasingly important role played by image memes in online communication, the research community does not have a good understanding of the elements that can have an effect in making a meme go viral on social media. In this paper, we investigate what visual elements influence the chances that an image meme will go viral, across three dimensions: composition, subjects, and target audience. Drawing from research in vision, psychology, marketing, and neuroscience, we develop a codebook to characterize image memes, and use it to annotate a set of 100 image memes collected from 4chan's Politically Incorrect Board (/pol/). We find that image memes that use a close-up scale and contain characters are more likely to go viral, and so are those including positive or negative emotions. Overall, our analysis sheds light on what indicators characterize viral visual content online, and set the basis for developing better techniques to create or moderate content that is more likely to catch the viewer's attention.

著者
Chen Ling
Boston University, Boston, Massachusetts, United States
Ihab AbuHilal
Binghamton University, Binghamton, New York, United States
Jeremy Blackburn
Emiliano De Cristofaro
Savvas Zannettou
Gianluca Stringhini
論文URL

https://doi.org/10.1145/3449155

動画

会議: CSCW2021

The 24th ACM Conference on Computer-Supported Cooperative Work and Social Computing

セッション: Social Media

Papers Room A
8 件の発表
2021-10-27 20:30:00
2021-10-27 22:00:00