Foresight Prediction Enhanced Live-Streaming Recommendation

TL;DR

This paper introduces a foresight prediction method for live-streaming recommendation that models content evolution to improve real-time content relevance and user engagement.

Contribution

It proposes a novel approach to predict future live-streaming content by semantic quantization and trend modeling, enhancing recommendation accuracy.

Findings

Improved user engagement during highlight moments

Enhanced recommendation relevance through foresight prediction

Effective in both offline and online experiments

Abstract

Live-streaming, as an emerging media enabling real-time interaction between authors and users, has attracted significant attention. Unlike the stable playback time of traditional TV live or the fixed content of short video, live-streaming, due to the dynamics of content and time, poses higher requirements for the recommendation algorithm of the platform - understanding the ever-changing content in real time and push it to users at the appropriate moment. Through analysis, we find that users have a better experience and express more positive behaviors during highlight moments of the live-streaming. Furthermore, since the model lacks access to future content during recommendation, yet user engagement depends on how well subsequent content aligns with their interests, an intuitive solution is to predict future live-streaming content. Therefore, we perform semantic quantization on live-streaming segments to obtain Semantic ids (Sid), encode the historical Sid sequence to capture the author's characteristics, and model Sid evolution trend to enable foresight prediction of future content. This foresight enhances the ranking model through refined features. Extensive offline and online experiments demonstrate the effectiveness of our method.