RPR

This is our experiment codes for the paper:

Review Polarity-wise Recommender.

Environment settings

• Python 3.7
• Tensorflow-GPU 1.14.0
• Numpy 1.19.5
• Pandas 1.1.3

File specification

• data_load.py : loads the raw json data in path ./raw_data.
• data_pro.py : processes the loaded data for model training, and the results are saved in path ./pro_data.
• word2vec_glove.py : processes the pre-trained word embeddings in path ./glove_embeddings for model training, and the results are saved in path ./pro_data.
• Model.py : implements the model framework of RPR.
• Model_train.py : integrates the training and testing process of RPR model.

Usage

• Execution sequence

  The execution sequence of codes is as follows : data_load.py--->data_pro.py--->word2vec_glove.py--->Model_train.py

• Execution results

  During the execution, the RPR performance on both training and testing sets will be printed after each optimization epoch:

  Epoch 0
  train_rmse, mae: 12.963 4.467
  test_loss 0.080, test_mse 12.945, test_rmse 3.598, test_mae 4.492
  
  Epoch 1
  train_rmse, mae: 12.960 4.444
  test_loss 0.079, test_mse 12.950, test_rmse 3.598, test_mae 4.490
  
  ...
  

  When the execution finished, the best performance on testing set will be printed:

  best_mse: 0.795
  best_rmse: 0.892
  best_mae: 0.652
  end