Recommender system enhanced with deep representations of customer reviews

Data sets

• Amazon Reviews Dataset: http://jmcauley.ucsd.edu/data/amazon/

• Goodreads Datasets: https://sites.google.com/eng.ucsd.edu/ucsdbookgraph

Project structure

• models/* - cdl and optimization algoruthms
• cdl_* - run models for godreads and amazon datasets
• data_preprocessing.ipynb - preprocess review texts with spaCy lib
• baselines_* - train baselines, using surprise lib
• helpful_stuff/* - utilities with output and evaluation methods
• graphs_* - variuos visualizations

CDL-based models

• cdl_conv_mf - CDL models with convolutional net for building reviews representations
• cld_sdae - general CDL model with SDAE for building reviews representations

Evaluation

• cdl_metrics - tests CDL model using MAE\RMSE and writes recall@M\precision@M to file

References

• Hao Wang, Naiyan Wang, and Dit-Yan Yeung. “Collaborative deep learning for recommender systems”. In:Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM. 2015, pages 1235–124

• Surprise library http://surpriselib.com/