HP-GMN: Graph Memory Networks for Heterophilous Graphs.
Junjie XuEnyan DaiXiang ZhangSuhang WangPublished in: ICDM (2022)
Keyphrases
- graph structures
- average degree
- real world graphs
- highly connected
- graph layout
- edge weights
- graph theory
- community discovery
- graph representation
- graph mining
- graph structure
- small world
- directed graph
- dynamic networks
- fully connected
- graph clustering
- weighted graph
- graph theoretic
- social networks
- social graphs
- random graphs
- graph matching
- labeled graphs
- complex networks
- protein interaction networks
- adjacency matrix
- betweenness centrality
- network analysis
- graph partitioning
- graph theoretical
- graph data
- bipartite graph
- graph model
- disk resident
- degree distribution
- graph databases
- graph properties
- graph kernels
- biological networks
- graph construction
- real world social networks
- graph search
- graph classification
- dense subgraphs
- densely connected
- small world networks
- reachability queries
- directed edges
- graph mining algorithms
- functional modules
- community detection
- maximum common subgraph
- dynamic graph
- series parallel
- shortest path
- network structure
- subgraph isomorphism
- connected components
- real world networks
- spanning tree
- connected graphs
- finding the shortest path
- community structure
- graph representations
- undirected graph
- power law
- graph isomorphism
- clustering coefficient
- maximum clique
- minimum spanning tree
- main memory
- structural pattern recognition
- scale free
- directed acyclic
- maximal cliques
- random walk
- directed acyclic graph
- polynomial time complexity
- evolving graphs