UltraGCN: Ultra Simplification of Graph Convolutional Networks for Recommendation.
Kelong MaoJieming ZhuXi XiaoBiao LuZhaowei WangXiuqiang HePublished in: CIKM (2021)
Keyphrases
- heterogeneous social networks
- directed graph
- fully connected
- edge weights
- highly connected
- average degree
- network structure
- network analysis
- recommender systems
- dynamic networks
- graph structures
- graph representation
- high speed
- social networks
- network size
- connected components
- multiresolution
- small world
- community discovery
- random walk
- protein interaction networks
- structured data
- bipartite graph
- real world graphs
- graphical representation
- biological networks
- directed edges
- directed acyclic graph
- weighted graph
- graph structure
- social graphs
- collaborative filtering
- deep learning
- citation networks
- random graphs
- social network analysis
- graph theoretic
- complex networks
- graph partitioning
- graph theory