In recommendation systems, high-quality user embeddings can capture subtle preferences, enable precise similarity calculations, and adapt to changing preferences over time to maintain relevance. The effectiveness of recommendation systems depends on the quality of user embedding. We propose to asynchronously learn high fidelity user embeddings for billions of users each day from sequence based multimodal user activities through a Transformer-like large scale feature learning module. The async learned user representations embeddings (ALURE) are further converted to user similarity graphs through graph learning and then combined with user realtime activities to retrieval highly related ads candidates for the ads delivery system. Our method shows significant gains in both offline and online experiments.
@article{arxiv.2406.05898,
title = {Async Learned User Embeddings for Ads Delivery Optimization},
author = {Mingwei Tang and Meng Liu and Hong Li and Junjie Yang and Chenglin Wei and Boyang Li and Dai Li and Rengan Xu and Yifan Xu and Zehua Zhang and Xiangyu Wang and Linfeng Liu and Yuelei Xie and Chengye Liu and Labib Fawaz and Li Li and Hongnan Wang and Bill Zhu and Sri Reddy},
journal= {arXiv preprint arXiv:2406.05898},
year = {2024}
}
Comments
Accepted by workshop on Multimodal Representation and Retrieval at SIGIR 2024, Washington DC