SocRipple: A Two-Stage Framework for Cold-Start Video Recommendations
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913982886969344 |
|---|---|
| author | Jaspal, Amit Dalwani, Kapil Ramineni, Ajantha |
| author_facet | Jaspal, Amit Dalwani, Kapil Ramineni, Ajantha |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_07241 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | SocRipple: A Two-Stage Framework for Cold-Start Video Recommendations Jaspal, Amit Dalwani, Kapil Ramineni, Ajantha Information Retrieval Artificial Intelligence 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. |
| title | SocRipple: A Two-Stage Framework for Cold-Start Video Recommendations |
| topic | Information Retrieval Artificial Intelligence |
| url | https://arxiv.org/abs/2508.07241 |