Researchers have introduced MaskRec, a novel architecture designed to enhance post-click conversion rate (CVR) prediction in large-scale advertising systems. This unified framework integrates heterogeneous feature interactions and multi-domain user behavior sequences within a single model. MaskRec utilizes a topology-masked attention mechanism called TopoMask to selectively manage information flow, allowing for more effective modeling of diverse data sources. Experiments conducted on the Tencent Advertising Algorithm Competition dataset demonstrated that MaskRec significantly outperforms existing baselines, highlighting its effectiveness for industrial CVR prediction. AI
IMPACT This architecture could improve the efficiency and accuracy of ad targeting systems by better modeling user behavior.
RANK_REASON The item is a research paper detailing a new model architecture. [lever_c_demoted from research: ic=1 ai=1.0]
Read on arXiv cs.IR (Information Retrieval) →
- alphaXiv
- CatalyzeX Code Finder for Papers
- Connected Papers
- CORE Recommender
- DagsHub
- Gotit.pub
- Hugging Face
- Influence Flower
- Litmaps
- MaskRec
- ScienceCast
- scite Smart Citations
- Tencent Advertising Algorithm Competition
- TopoMask
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →