To tailor the contrastive learning task to the CVR prediction problem, we propose embedding masking (EM), rather than feature masking, to create two views of augmented samples.
For existing old ads, GMEs first build a graph to connect them with new ads, and then adaptively distill useful information.
Our study is based on UC Toutiao (a news feed service integrated with the UC Browser App, serving hundreds of millions of users daily), where the source domain is the news and the target domain is the ad.
Both offline and online experiments demonstrate the effectiveness of MA-DNN for practical CTR prediction services.
These subnets model the user-ad, ad-ad and feature-CTR relationship respectively.
Ranked #2 on Click-Through Rate Prediction on Avito
The intuitions are that ads shown together may influence each other, clicked ads reflect a user's preferences, and unclicked ads may indicate what a user dislikes to certain extent.
Ranked #1 on Click-Through Rate Prediction on Avito