FOCAL fuses unconstrained coverage attention and meta-path anchoring attention to improve multi-label classification on heterogeneous graphs by resolving semantic dilution versus coverage constraint trade-offs.
The world wide web conference , pages=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
FOCAL-Attention for Heterogeneous Multi-Label Prediction
FOCAL fuses unconstrained coverage attention and meta-path anchoring attention to improve multi-label classification on heterogeneous graphs by resolving semantic dilution versus coverage constraint trade-offs.