Code-for-AEGCN-Glove

Code for applied sciences paper "Targeted Sentiment Classification Based on Attentional Encoding and Graph Convolutional Networks"

Model

Overall architecture of AEGCN

Requirements

• Python >= 3.6
• pytorch >= 1.0
• SpaCy 2.0.18
• numpy 1.15.4
• sklearn

Usage

• First install SpaCy package and language models with

pip install spacy

and

python -m spacy download en

• Download pretrained GloVe embeddings here and put glove.840B.300d.txt into glove/.
• Train with [train.py]

Citation

If you find this repository is helpful to you, please kindly cite our paper and star this repo

@article{xiao2020targeted,
    title={Targeted Sentiment Classification Based on Attentional Encoding and Graph Convolutional Networks},
    author={Xiao, Luwei and Hu, Xiaohui and Chen, Yinong and Xue, Yun and Gu, Donghong and Chen, Bingliang and Zhang, Tao},
    journal={Applied Sciences},
    volume={10},
    number={3},
    pages={957},
    year={2020},
    publisher={Multidisciplinary Digital Publishing Institute}
}

Acknowledgement

This work is mainly based on the repositories of ASGCN and ABSA-PyTorch.Thanks a million to anyone who assists with this work and their valuable devotion!