Relational-Graph-Attention-Network-for-Aspect-based-Sentiment-Analysis

This repo contains the PyTorch implementaion for the paper Relational Graph Attention Network for Aspect-based Sentiment Analysis.

For any questions about the implementation, plaese email shenwzh3@mail2.sysu.edu.cn.

Requirements

• Python 3.6.8
• PyTorch 1.2.0
• Transformers https://github.com/huggingface/transformers
• CUDA 9.0

Perparation

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.

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.

Preprocess

The preprocess codes are in data_preprocess_semeval.py and data_preprocess_twitter.py. However we already provided the preprocessed datasets with dependency parcing results in ./data/, so you can skip preprocess.

Training

Run: ./run.sh