SMTPD: A New Benchmark for Temporal Prediction of Social Media Popularity

2025-03-06

AI Summary

Existing social media popularity prediction research largely neglects temporal alignment modeling, hindering accurate characterization of the time-coupled relationship between early signals and subsequent popularity evolution. To address this, we propose SMTPD—the first benchmark for temporally aligned social media popularity prediction—built upon multilingual, multimodal YouTube data. SMTPD systematically incorporates temporal alignment mechanisms to uncover the temporal synergy between early and long-term popularity dynamics. We further introduce a novel evaluation paradigm balancing timeliness and interpretability, integrating multimodal features, temporally aligned modeling, and cross-lingual popularity normalization, alongside a lightweight dynamic prediction baseline. Experiments demonstrate that temporal alignment reduces mean absolute error (MAE) by 18.7%. The benchmark dataset, evaluation framework, and code are publicly released to advance standardized, reproducible research in this domain.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Data Mining & Knowledge Management: Mining of Spatial, Temporal or Spatio-Temporal Data
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)

Application Category

• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Social media popularity prediction task aims to predict the popularity of posts on social media platforms, which has a positive driving effect on application scenarios such as content optimization, digital marketing and online advertising. Though many studies have made significant progress, few of them pay much attention to the integration between popularity prediction with temporal alignment. In this paper, with exploring YouTube's multilingual and multi-modal content, we construct a new social media temporal popularity prediction benchmark, namely SMTPD, and suggest a baseline framework for temporal popularity prediction. Through data analysis and experiments, we verify that temporal alignment and early popularity play crucial roles in social media popularity prediction for not only deepening the understanding of temporal dynamics of popularity in social media but also offering a suggestion about developing more effective prediction models in this field. Code is available at https://github.com/zhuwei321/SMTPD.

Problem

• Predict social media post popularity with temporal alignment.
• Develop a benchmark for temporal popularity prediction on YouTube.
• Explore early popularity's role in effective prediction models.

Innovation

• Developed SMTPD benchmark for temporal popularity prediction
• Integrated temporal alignment with popularity prediction models
• Utilized YouTube's multilingual and multi-modal content data