Abstract:Current state-of-the-art Online Video Services (OVSs) need to simultaneously serve hundreds of millions of users at peak time. To improve service efficiency, the caching strategies of their Content Delivery Networks (CDNs) usually work with video segments of a couple of seconds in length, which poses great challenges in popular prediction of video segments. Popularity prediction of entire videos is generally based on the correlation of early views and future views. However, for the popularity prediction of video segments, the first and the rest segments of a video should be distinguished first because users usually start watching a video from its first segment, but could skip any of the rest ones. Towards this end, we propose a novel method for video segment popularity prediction, in which, the popularity of videos' first and rest segments are predicted with different models. For the first segments, the popularity is predicted like entire videos by using a Multi-Linear Regression (MLR) Model based on past popularity. And for the rest segments, as the jumping viewing behavior has largely weaken the indicating power of their early views on future views, we predict their popularity based on the recent viewed previous segments. Specially, considering the real-time requirement of CDNs caching, it is infeasible to investigate the viewed segments sequences of users with sequence models, hence the prediction model is built on statistic level. Moreover, to maintain the performance in online prediction, we equip the two prediction models with online learning capability so as to timely capture the changes in popularity. For promptly service, the job is fulfilled with an online compensator instead of retraining the two prediction models with revealed target of past predictions. The online compensator estimates the prediction errors of the two models through a LSTM network based on their recent errors and the popularity data corresponded to recent tasks. Through extensive experiments on real-world data of iQIYI, one of the most popular OVS providers in China, we have demonstrated the promise of the proposed method.

A Novel Time Series Approach for Predicting the Long-Term Popularity of Online Videos

Modeling Dynamics of Online Video Popularity

A Real-Time Method to Predict Social Media Popularity

A First View on Mobile Video Popularity As Time Series.

Forecasting Popularity of Videos using Social Media

Predicting the Top-N Popular Videos Via a Cross-Domain Hybrid Model

Predicting the Popularity of Viral Topics Based on Time Series Forecasting

Online Popularity Prediction of Video Segments Towards More Efficient Content Delivery Networks

On popularity prediction of videos shared in online social networks.

Online Prediction of Video Popularity in OVSs: A Video Age-Sensitive Model with Beyond Views Features

Online Video Popularity Regression Prediction Model with Multichannel Dynamic Scheduling Based on User Behavior

Predicting the Popularity of Online Serials with Autoregressive Models.

Lifetime popularity prediction for online videos

Modeling and Quantifying the Forces Driving Online Video Popularity Evolution

An Intelligent Video Tag Recommendation Method for Improving Video Popularity in Mobile Computing Environment

A Video Popularity Prediction Scheme with Attention-Based LSTM and Feature Embedding.

Adaptive time series prediction and recommendation

User Behavior Analysis and Video Popularity Prediction on a Large-Scale VoD System

Forecasting Initial Popularity of Just-Uploaded User-Generated Videos.

Hybrid Machine Learning Approach to Popularity Prediction of Newly Released Contents for Online Video Streaming Service

Predicting the Online Performance of Video Service Providers on the Internet