Papers with Long-Tail Recommendation
MotiR: Motivation-aware Retrieval for Long-Tail Recommendation (2025.acl-industry)
Copied to clipboard
| Challenge: | Existing methods for retrieval of recommendation systems rely on collaborative filtering signals and lacks similarity for long-tail items. |
| Approach: | They propose a Motivation-aware Retrieval for Long-Tail Recommendation that integrates purchase motivations with traditional item features to capture similarity among long-tail items. |
| Outcome: | The proposed model captures similarity between long-tail items while maintaining collaborative filtering advantages for popular items. |