-OS Ubuntu 16.04.6 LTS
-python 3.7.6
-GPU: NVIDIA GeForce RTX 3090(Driver version:460.56, CUDA Version:11.2)
-Semeval14 laptop
-Semeval14 rest
-Twitter
-ds_name: name of the dataset, options:14semeval_rest 14semeval_laptop Twitter
-bs: batch_size
-learning_rate: default 0.001
-n_epoch: number of epoch for training, default 20
-test: whether test model or not, default 0
-run.sh or python ./main_total.py
-Half of our code are based on DM-GCN(Pang et.al, ACL2021) and KGAN(Zhong et.al, arXiv preprint)