A weighted GAT extension of the region2vec graph embedding method jointly uses node attributes, mobility flows, and geographic adjacency to detect spatial network communities, and outperforms baselines on a composite score in Wisconsin tract experiments.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SI 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
GeoAI-Enhanced Community Detection on Spatial Networks with Graph Deep Learning
A weighted GAT extension of the region2vec graph embedding method jointly uses node attributes, mobility flows, and geographic adjacency to detect spatial network communities, and outperforms baselines on a composite score in Wisconsin tract experiments.