MGA-HHN constructs meta-path based heterogeneous hypergraphs and applies node-level and hyperedge-level attention, reporting improved node classification, clustering, and visualization on three standard heterogeneous graph benchmarks.
Heterogeneous network representation learning,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Multi-Granular Attention based Heterogeneous Hypergraph Neural Network
MGA-HHN constructs meta-path based heterogeneous hypergraphs and applies node-level and hyperedge-level attention, reporting improved node classification, clustering, and visualization on three standard heterogeneous graph benchmarks.