Dissecting the Meme Magic: Understanding Indicators of Virality in Image Memes (2101.06535v1)

Published 16 Jan 2021 in cs.HC, cs.CY, and cs.SI

Abstract: Despite the increasingly important role played by image memes, we do not yet have a solid understanding of the elements that might make a meme go viral on social media. In this paper, we investigate what visual elements distinguish image memes that are highly viral on social media from those that do not get re-shared, across three dimensions: composition, subjects, and target audience. Drawing from research in art theory, 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/). On the one hand, we find that highly viral memes are more likely to use a close-up scale, contain characters, and include positive or negative emotions. On the other hand, image memes that do not present a clear subject the viewer can focus attention on, or that include long text are not likely to be re-shared by users. We train machine learning models to distinguish between image memes that are likely to go viral and those that are unlikely to be re-shared, obtaining an AUC of 0.866 on our dataset. We also show that the indicators of virality identified by our model can help characterize the most viral memes posted on mainstream online social networks too, as our classifiers are able to predict 19 out of the 20 most popular image memes posted on Twitter and Reddit between 2016 and 2018. Overall, our analysis sheds light on what indicators characterize viral and non-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.

View on arXiv

Authors (6)

Chen Ling (65 papers)
Ihab AbuHilal (1 paper)
Jeremy Blackburn (76 papers)
Emiliano De Cristofaro (117 papers)
Savvas Zannettou (55 papers)
Gianluca Stringhini (77 papers)

Citations (49)

View on Semantic Scholar

Summary

Analysis of Meme Virality Indicators

The paper "Dissecting the Meme Magic: Understanding Indicators of Virality in Image Memes" provides an analytical paper of the elements that influence the virality of image memes across social media platforms. The authors systematically explore the factors that contribute to an image meme's likelihood of being re-shared widely, specifically dissecting components related to composition, subjects, and target audience.

Methodology and Findings

The researchers employ a mixed-methods approach, starting with the development of a codebook that draws on interdisciplinary research from art theory, psychology, neuroscience, and marketing. They annotate a dataset of 100 image memes from 4chan's /pol/ board, categorizing them as either viral or non-viral based on their prevalence. The paper relies on machine learning models trained to identify memes likely to go viral, achieving an impressive AUC of 0.866 on the dataset. Notably, these models accurately predict 95% of the top viral memes on Twitter and Reddit, thereby demonstrating their robustness.

Key findings highlight that image memes are more likely to go viral if they:

Use a close-up scale, allowing focused engagement with the subject.
Depict characters over objects, thereby attracting viewer attention.
Feature emotions—whether positive or negative—that resonate strongly.
Present a well-balanced composition, facilitating visual appeal.

In contrast, memes with long text or those lacking a clear subject are unlikely to be re-shared extensively. The results align with existing theories in visual art and psychology, underscoring how factors such as composition and subject matter significantly affect meme virality.

Implications and Future Prospects

This research provides valuable insights applicable in domains such as marketing, online communication strategies, and social media moderation. If leveraged wisely, it can aid in crafting engaging visuals for campaigns or improving content moderation on social platforms by pinpointing potentially harmful viral content.

Additionally, it opens avenues for further exploration into community-specific meme characteristics—what resonates within niche groups versus broader audiences—and how meme culture evolves over time. Such studies could further refine machine learning models to predict not just virality but also potential human reception and interaction patterns across different cultural settings or platforms.

Conclusion

The paper aptly dissects meme virality, establishing a foundation on which future studies may build. While the research highlights intriguing patterns that dictate meme popularity, it also raises broader questions about digital content creation and dissemination practices. The paper's methodological rigor and its promise for application in various fields position it as an essential reference for researchers interested in social media dynamics and cultural production. The findings are substantial, suggesting that both the aesthetic quality of memes and the emotional tenor they convey are crucial determinants of their ability to capture public imagination and propagate virally across the digital landscape.

Related Papers

Find Related Papers

YouTube

Show All Videos