LLM-Alignment Live-Streaming Recommendation
By: Yueyang Liu , Jiangxia Cao , Shen Wang and more
Potential Business Impact:
Shows you live videos you'll like best.
In recent years, integrated short-video and live-streaming platforms have gained massive global adoption, offering dynamic content creation and consumption. Unlike pre-recorded short videos, live-streaming enables real-time interaction between authors and users, fostering deeper engagement. However, this dynamic nature introduces a critical challenge for recommendation systems (RecSys): the same live-streaming vastly different experiences depending on when a user watching. To optimize recommendations, a RecSys must accurately interpret the real-time semantics of live content and align them with user preferences.
Similar Papers
Foresight Prediction Enhanced Live-Streaming Recommendation
Information Retrieval
Predicts future live video to show you what you'll like.
User Feedback Alignment for LLM-powered Exploration in Large-scale Recommendation Systems
Information Retrieval
Finds new videos you'll like, not just favorites.
Dynamic Content Moderation in Livestreams: Combining Supervised Classification with MLLM-Boosted Similarity Matching
CV and Pattern Recognition
Stops bad videos from being watched by more people.