Doc-GCN: Heterogeneous Graph Convolutional Networks for Document Layout Analysis.
Siwen LuoYihao DingSiqu LongSoyeon Caren HanJosiah PoonPublished in: CoRR (2022)
Keyphrases
- heterogeneous networks
- average degree
- document images
- graph structures
- random walk
- document collections
- edge weights
- graph theory
- dynamic networks
- graph representation
- social networks
- protein interaction networks
- small world
- weighted graph
- directed graph
- information retrieval
- web documents
- directed edges
- community discovery
- network size
- fully connected
- keywords
- information retrieval systems
- structured data
- text documents
- document retrieval
- graph mining
- random graphs
- deep learning
- graphical representation
- multiple types
- metadata
- betweenness centrality
- social graphs
- connected components
- graph layout
- bipartite graph