SocRipple: A Two-Stage Framework for Cold-Start Video Recommendations
This addresses the challenge of distributing new items effectively in social graph-based platforms, offering a domain-specific solution that is incremental in nature.
The paper tackles the cold-start problem in recommender systems for new items lacking interaction history by proposing SocRipple, a two-stage framework that leverages social connections and early engagement signals, resulting in a 36% boost in cold-start item distribution while maintaining user engagement rates.
Most industry scale recommender systems face critical cold start challenges new items lack interaction history, making it difficult to distribute them in a personalized manner. Standard collaborative filtering models underperform due to sparse engagement signals, while content only approaches lack user specific relevance. We propose SocRipple, a novel two stage retrieval framework tailored for coldstart item distribution in social graph based platforms. Stage 1 leverages the creators social connections for targeted initial exposure. Stage 2 builds on early engagement signals and stable user embeddings learned from historical interactions to "ripple" outwards via K Nearest Neighbor (KNN) search. Large scale experiments on a major video platform show that SocRipple boosts cold start item distribution by +36% while maintaining user engagement rate on cold start items, effectively balancing new item exposure with personalized recommendations.