Delving Deep into Engagement Prediction of Short Videos

TL;DR

This paper introduces a new dataset and metrics for predicting engagement of short videos, revealing that traditional quality scores are not reliable indicators of engagement, and explores multi-modal features for improved prediction.

Contribution

It provides a large-scale real-world dataset and proposes novel engagement metrics, advancing the understanding of short video engagement prediction.

Findings

Mean Opinion Scores do not correlate well with engagement

Proposed NAWP and ECR metrics effectively describe engagement levels

Multi-modal features improve engagement prediction accuracy

Abstract

Understanding and modeling the popularity of User Generated Content (UGC) short videos on social media platforms presents a critical challenge with broad implications for content creators and recommendation systems. This study delves deep into the intricacies of predicting engagement for newly published videos with limited user interactions. Surprisingly, our findings reveal that Mean Opinion Scores from previous video quality assessment datasets do not strongly correlate with video engagement levels. To address this, we introduce a substantial dataset comprising 90,000 real-world UGC short videos from Snapchat. Rather than relying on view count, average watch time, or rate of likes, we propose two metrics: normalized average watch percentage (NAWP) and engagement continuation rate (ECR) to describe the engagement levels of short videos. Comprehensive multi-modal features, including visual content, background music, and text data, are investigated to enhance engagement prediction. With the proposed dataset and two key metrics, our method demonstrates its ability to predict engagements of short videos purely from video content.