HAConvGNN: Hierarchical Attention Based Convolutional Graph Neural Network for Code Documentation Generation in Jupyter Notebooks.
Xuye LiuDakuo WangApril Yi WangYufang HouLingfei WuPublished in: EMNLP (Findings) (2021)
Keyphrases
- neural network
- network architecture
- graph representation
- graph structure
- neural network model
- random walk
- back propagation
- neural network is trained
- graph theory
- source code
- directed graph
- weighted graph
- graph model
- hierarchical structure
- feed forward
- multi layer
- artificial neural networks
- restricted boltzmann machine
- generation process
- directed acyclic graph
- fuzzy logic
- deep learning
- structured data
- fuzzy artmap
- deep belief networks
- bp neural network
- graph theoretic
- graph databases
- self organizing maps
- fault diagnosis
- image classification
- pattern recognition
- multiscale