UltraGCN: Ultra Simplification of Graph Convolutional Networks for Recommendation.
Kelong MaoJieming ZhuXi XiaoBiao LuZhaowei WangXiuqiang HePublished in: CoRR (2021)
Keyphrases
- heterogeneous social networks
- average degree
- small world
- graph theory
- collaborative filtering
- complex networks
- directed graph
- fully connected
- edge weights
- graph representation
- recommender systems
- graph matching
- random walk
- network analysis
- dynamic networks
- citation networks
- high speed
- graph theoretic
- betweenness centrality
- real world graphs
- degree distribution
- network size
- directed edges
- graph model
- graph structures
- heterogeneous networks
- weighted graph
- graph structure
- social networks
- deep learning
- neural network
- bipartite graph
- connected components
- structured data
- preprocessing