sentiment-review-summary

Code for our COLING-2020 paper "Making the Best Use of Review Summary for Sentiment Analysis"

Before running the code

• Our experiments were conducted under pytorch==1.0.1 and cudatoolkit==9.0, with python==3.6.
• Some other required packages: tensorboardX, pickle, nltk, numpy.
• SNAP Amazon Review dataset can be obtained from this url. (We cannot provide the data in our repo due to copyright issue. )
• The paths of the dataset and the embedding file (we use GloVe in our experiments) in the config file need to be filled in before running the code.

Run the code

python main.py --config config_toy

• /param/config_toy.py is just an example config file. You may create your own config file.

Cite

If you find our code useful, please consider citing our paper:

@inproceedings{yang-etal-2020-making,
    title = "Making the Best Use of Review Summary for Sentiment Analysis",
    author = "Yang, Sen  and
      Cui, Leyang  and
      Xie, Jun  and
      Zhang, Yue",
    booktitle = "Proceedings of the 28th International Conference on Computational Linguistics",
    month = dec,
    year = "2020",
    address = "Barcelona, Spain (Online)",
    publisher = "International Committee on Computational Linguistics",
    url = "https://www.aclweb.org/anthology/2020.coling-main.15",
    pages = "173--184",
}