DTH-GNN, a graph neural network with temporal edge decomposition, heterogeneous transformers, and contrastive learning, is claimed to improve ad feedback prediction AUC by 8.2% on the Coveo dataset, but evidence is incomplete.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Analysis of Anonymous User Interaction Relationships and Prediction of Advertising Feedback Based on Graph Neural Network
DTH-GNN, a graph neural network with temporal edge decomposition, heterogeneous transformers, and contrastive learning, is claimed to improve ad feedback prediction AUC by 8.2% on the Coveo dataset, but evidence is incomplete.