/ADMN

A tensorflow implementation for "Attention-based Adaptive Memory Network for Recommendation with Review and Rating"

Primary LanguagePythonMIT LicenseMIT

ADMN

A tensorflow implementation for the paper "Attention-based Dynamic Memory Network for Recommendation with Review and Rating"

Enviorments and Catelog

Enviorments

  • Python 3.5
  • Tensorflow 1.7
  • Keras 2.0.9
  • numpy 1.18
  • pandas 1.0.5

Catelog

  • Data: dataset
  • conf: Profile about hyperparameter
  • Model: ADMN model
  • pro_data: pre-processing model
  • Result: save results
  • temp: temporary file
  • main.py: Main program

python code

Examples to run the code can be found in the main.py. Run main.py

python main.py

noted: If you want to modify hyperparameters, such as datasets, you need to modify the profile(conf/conf.py)