ME-GCN: Multi-dimensional Edge-Embedded Graph Convolutional Networks for Semi-supervised Text Classification.
Kunze WangSoyeon Caren HanSiqu LongJosiah PoonPublished in: CoRR (2022)
Keyphrases
- text classification
- multi dimensional
- semi supervised
- edge weights
- labeled data
- weighted graph
- unlabeled data
- co training
- graph construction
- semi supervised classification
- label propagation
- semi supervised learning
- graph theory
- labeled and unlabeled data
- text categorization
- undirected graph
- random walk
- densely connected
- bipartite graph
- vertex set
- disjoint paths
- text mining
- complex networks
- feature selection
- bag of words
- machine learning
- pairwise
- structured data
- supervised learning
- multi label
- data cleaning
- small world
- unsupervised learning
- network structure
- fully connected
- edge detection
- n gram
- graph structure
- semantic features
- directed graph
- graph structures
- multiple dimensions
- active learning
- network analysis
- shortest path
- deep learning
- spanning tree
- graph mining
- strongly connected