Skip to content

Latest commit

 

History

History

RGAT

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
LICENSE
LICENSE
 
 
README.md
README.md
 
 
datasets.py
datasets.py
 
 
model.py
model.py
 
 
model_gcn.py
model_gcn.py
 
 
model_utils.py
model_utils.py
 
 
run.py
run.py
 
 
trainer.py
trainer.py
 
 
tree.py
tree.py
 
 

README.md

This is modified from the RGAT, which is the source code of the paper Relational Graph Attention Network for Aspect-based Sentiment Analysis.

We made neccessary changes. We believe all the changes are under the MIT License permission.

Usage

  1. For Glove Embedding

First, download and unzip GloVe vectors(glove.840B.300d.zip) from https://nlp.stanford.edu/projects/glove/. Then change the value of parameter --glove_dir to the directory of the word vector file.

  1. For BERT Embedding

Download the pytorch version pre-trained bert-base-uncased model and vocabulary from the link provided by huggingface. Then change the value of parameter --bert_model_dir to the directory of the bert model.

  1. Train with command
python run.py --highway --dataset_name the/dataset/path
  • the --dataset arguement should be the path of dataset.

Notes

  1. RGAT model take input files in different format, which should be generated after the running of the Perturbed-Masking code.

  2. The --dataset should be the data path (/user/project/dataset/Resaurant) rather than the dataset name:

/user/project/dataset/
(--dataset) |---Restaurant
            |------Train
            |------Test
(--dataset) |---Laptop
            |------Train
            |------Test
(--dataset) |---fr
            |------Train
            |------Test
...