A post-hoc rating compression plus a high-confidence negative-sample loss reduces overconfidence and slightly improves recommendation accuracy in GNN-based recommender systems.
Adaptive denoising graph con- trastive learning with memory graph attention for recommendation,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.IR 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
Conf-GNNRec: Quantifying and Calibrating the Prediction Confidence for GNN-based Recommendation Methods
A post-hoc rating compression plus a high-confidence negative-sample loss reduces overconfidence and slightly improves recommendation accuracy in GNN-based recommender systems.