BilinearReview_journal

Personalized Sentiment Classification of Customer Reviews via an Interactive Attributes Attention Model

Main Requirements

transformers
yaml
easydict
torchtext
spacy

all datasets should be downloaded and unzipped in the directory corpus in the orginal codes
- Original IMDB, Yelp datasets are available at here and Amazon dataset at here
- zipped datasets are also downloaded at here (提取码：luck)
running a model over IMDB datasets

python run.py --run train --dataset imdb --model bilinear --gpu 0,1

All training Hyper-Parameters are set in cfgs/config.py and model parameters in cfgs/bilinear_model.yml.

Name		Name	Last commit message	Last commit date
Latest commit History 5 Commits
.idea		.idea
cfgs		cfgs
common		common
datasets		datasets
logs		logs
models_v2		models_v2
README.md		README.md
__init__.py		__init__.py
get_save_vectors.py		get_save_vectors.py
refine_bert.py		refine_bert.py
run.py		run.py