Predicting the Popularity of Online Videos via Deep Neural Networks

TL;DR

This paper introduces a deep learning approach combining multi-task learning and relation networks to improve the accuracy of predicting online video popularity, addressing the complexity and competition involved.

Contribution

It presents a novel model integrating MTL and RN modules for more accurate popularity prediction of online videos, especially in multi-competitor scenarios.

Findings

Significant increase in prediction accuracy for total view counts.

Effective modeling of relationships among multiple competitors.

Reduction of over-fitting through multi-task learning.

Abstract

Predicting the popularity of online videos is important for video streaming content providers. This is a challenging problem because of the following two reasons. First, the problem is both "wide" and "deep". That is, it not only depends on a wide range of features, but also be highly non-linear and complex. Second, multiple competitors may be involved. In this paper, we propose a general prediction model using the multi-task learning (MTL) module and the relation network (RN) module, where MTL can reduce over-fitting and RN can model the relations of multiple competitors. Experimental results show that our proposed approach significantly increases the accuracy on predicting the total view counts of TV series with RN and MTL modules.