Predicting Session Length in Media Streaming

TL;DR

This paper introduces a novel approach to predict session lengths in mobile music streaming using survival analysis and gradient boosted trees, demonstrating improved accuracy over baselines in real-world data.

Contribution

It is the first to analyze session length prediction in a mobile streaming context using survival analysis and machine learning techniques.

Findings

Survival analysis reveals significant differences in session length distributions among users.

Gradient boosted trees effectively predict session length using initial session data.

Proposed method outperforms baseline models on real-world streaming data.

Abstract

Session length is a very important aspect in determining a user's satisfaction with a media streaming service. Being able to predict how long a session will last can be of great use for various downstream tasks, such as recommendations and ad scheduling. Most of the related literature on user interaction duration has focused on dwell time for websites, usually in the context of approximating post-click satisfaction either in search results, or display ads. In this work we present the first analysis of session length in a mobile-focused online service, using a real world data-set from a major music streaming service. We use survival analysis techniques to show that the characteristics of the length distributions can differ significantly between users, and use gradient boosted trees with appropriate objectives to predict the length of a session using only information available at its beginning. Our evaluation on real world data illustrates that our proposed technique outperforms the considered baseline.