Researchers have developed a new user foundation model designed for the open web, addressing the challenges of fragmented and non-persistent user identities. This model utilizes self-supervised learning on user browsing histories, applying a Transformer encoder pre-trained with masked language modeling and a sequence-level contrastive objective. The approach has shown significant improvements in downstream production tasks, including a 2.13% increase in CTR and a 1.13% decrease in eCPC during a 7-day A/B test. AI
IMPACT This research could improve personalization and ad targeting on the open web by better utilizing fragmented user data.
RANK_REASON The cluster contains an academic paper detailing a new model and its performance evaluation. [lever_c_demoted from research: ic=1 ai=1.0]
- alphaXiv
- arXiv
- CatalyzeX
- Connected Papers
- CORE Recommender
- DagsHub
- Gotit.pub
- Hugging Face
- IArxiv Recommender
- Litmaps
- ScienceCast
- scite Smart Citations
- Transformer++
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
