MOMENTUM MEETS VIRALITY: A NOVEL METRIC FOR UNMASKING SOCIAL BIAS IN VIRAL TWEETS
Keywords: Social bias, Tweet virality, ViralTweetScore, Hindi Tweets, Tweets
TL;DR: We introduce a novel metric, ViralTweet Score (VTS), inspired by momentum principles, to better predict tweet virality, compare it with existing metrics, and highlight how social biases influence virality, using a dataset of 88.8k Hindi tweets.
Abstract: Predicting which social media posts will go viral is a critical but complex task in the field of computational social science. Previous studies have utilized various measures to forecast the virality of tweets or Facebook posts, but these approaches exhibit limitations, particularly in the absence of a virality metric that specifically considers social biases. In this paper, we test existing metrics and introduce a new metric, $\textbf{ViralTweet Score (VTS)}$, inspired by principles of momentum from physics to better predict a tweet's virality given that it consists of social biases. We compare this new metric with others, highlighting the advantages and disadvantages of each of them as a virality measurement metric. We release the $\textbf{ViralTweets Dataset}$ with $\mathbf{88.8k}$ Hindi tweets and corresponding virality labels based on our VTS metric. We also show how social biases in posts can influence their potential to go viral. We test our hypothesis that VTS is a better metric using two methodologies and we show how VTS achieves an F1 score of 0.87 based on pairwise evaluation methodology and an overall F1 score of 0.58 based on our clustering-based verification methodology. Our work offers a novel metric for understanding tweet virality for biased tweets and opens the door for more equitable and effective social media analytics by considering the role of social biases in virality.
Primary Area: alignment, fairness, safety, privacy, and societal considerations
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.
Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 14245
Loading