Source code of the URLGraph++

The graph

-VocabularyGraph.py:the vocabulary-level graph we build

-CharacterGraph.py:the character-level graph we build

The model

-vgcn-bert.py:The proposed model we use

-vgcn-bertNoChar.py:The model that move away CharCNN

-vgcn-bertNoDynMM.py:The model that move away DynMM

The dataset

-custom_dataset.py:To split the dataset into validate and train and pre-process the data

To run the code

-MultipleFusion.py:The code that train the proposed model and get the metrics in 1,6,12 layers

-NoFusion.py:The code that train the vgcn-bertNoChar and get the metrics in 1,6,12 layers

-FusionWithoutDynMM:The code that train the vgcn-bertNoDynMM and get the metrics in 1,6,12 layers